Liu2016.pdf - Edinburgh Research Archive

This thesis has been submitted in fulfilment of the requirements for a postgraduate degree

(e.g. PhD, MPhil, DClinPsychol) at the University of Edinburgh. Please note the following

terms and conditions of use:

This work is protected by copyright and other intellectual property rights, which are

retained by the thesis author, unless otherwise stated.

A copy can be downloaded for personal non-commercial research or study, without

prior permission or charge.

This thesis cannot be reproduced or quoted extensively from without first obtaining

permission in writing from the author.

The content must not be changed in any way or sold commercially in any format or

medium without the formal permission of the author.

When referring to this work, full bibliographic details including the author, title,

awarding institution and date of the thesis must be given.

PhD Psychology – The University of Edinburgh – 2015

Understanding Optimism Caimei Liu

Contents

Acknowledgements ..................................................................................................... 6

Declaration .................................................................................................................. 7

Published works ......................................................................................................... 8

Abstract ....................................................................................................................... 9

Chapter 1: What is optimism? .................................................................................. 1

1.1 Origins and concepts of optimism................................................................ 1

1.2 Explanatory Style ......................................................................................... 4

1.2.1 Historical Development of models of explanatory Style ......................... 4

1.2.2 Measures of explanatory style .................................................................. 7

1.2.3 Stability and heritability of explanatory style ........................................ 11

1.2.4 Self-serving attributional bias and optimistic explanatory style ............ 12

1.2.5 Explanatory style, hopelessness, and depression ................................... 15

1.3 Dispositional Optimism ............................................................................. 16

1.3.1 Historical development of models of dispositional optimism ............... 16

1.3.2 Measures of dispositional optimism ...................................................... 19

1.3.3 Stability and heritability of dispositional optimism ............................... 20

1.4 Benefits of Optimism ................................................................................. 23

1.4.1 Optimism and physical well-being ........................................................ 24

1.4.2 Optimism and psychological well-being ................................................ 25

1.4.3 Optimism, resources, and success .......................................................... 29

1.4.4 Optimism interventions included in positive psychology interventions 31

1.4.5 Underlying mechanism: optimism and coping ...................................... 33

1.5 Outline of the current research ................................................................... 36

1.5.1 Optimism in positive psychology .......................................................... 36

1.5.2 Part I measurement and concepts of optimism ...................................... 38

1.5.3 Part II optimism interventions ................................................................ 40

1.5.4 Measures ................................................................................................ 41

1.5.5 Participants ............................................................................................. 47

Chapter 2: The psychometric construct of optimism ........................................... 50

2.1 The psychometric construct of the ASQ .................................................... 50

2.1.1 Myths about attributional style .............................................................. 50

2.1.2 Samples and instruments ........................................................................ 58

2.1.3 Testing models of causal attributions for positive and negative events . 59

2.1.4 Structural equation modeling ................................................................. 60

2.1.5 Replication final ASQ model ................................................................. 64

2.1.6 Schematic model of attributional style ................................................... 67

2.2 Separating optimism and pessimism .......................................................... 70

2.2.1 Previous understanding of dispositional optimism ................................ 70

2.2.2 Two-factor structure of the LOT ............................................................ 75

2.2.3 What we should know about dispositional optimism ............................ 80

Chapter 3: Optimism and personality ................................................................... 82

3.1 Is optimism a personality thing? ................................................................ 82

3.2 Methods ...................................................................................................... 91

3.3 Results ........................................................................................................ 92

3.4 Optimism and the Five-Factor Model of personality ............................... 112

Chapter 4: Optimism and psychological well-being ........................................... 117

4.1 Optimism and two approaches of well-being ........................................... 117

4.2 Samples and instruments .......................................................................... 123

4.3 Results ...................................................................................................... 125

4.4 Positive relationship between optimism and psychological well-being .. 133

Chapter 5: Cultural influence on optimism ......................................................... 136

5.1 Cultural issues: from the West to the East ............................................... 136

5.2 Prior studies investigating cultural differences in optimism .................... 137

5.3 The present study ..................................................................................... 141

5.3.1 Method ................................................................................................. 143

5.3.2 Results .................................................................................................. 144

5.3.3 Are Chinese people more optimistic than British people? ................... 155

Chapter 6: Extending thoughts on attributional bias ......................................... 158

6.1 What we know and what we don’t know about attributional bias ........... 158

6.2 Attributional evaluation system and possible attributional models ......... 162

6.3 Psychometric structure of the ASQ-Other ............................................... 166

6.4 Study 1: testing attributional models using ASQ and ASQ-Other ........... 172

6.5 Study 2: testing event-focused attibutional style using ASQ-General ..... 178

6.6 Attributional biases in reality ................................................................... 181

Chapter 7: Depression, positive psychology and optimism interventions ........ 184

7.1 Traditional treatments for depression ....................................................... 185

7.2 Rising of positive psychology interventions ............................................ 187

7.3 Optimism and depression ......................................................................... 188

7.3.1 Attributional style in depression .......................................................... 189

7.3.2 Dispositional optimism and depression ............................................... 195

7.4 How to manipulate optimism? ................................................................. 197

7.5 Empirical studies of optimism interventions ............................................ 200

7.5.1 Optimism interventions in nonclinical samples ................................... 201

7.5.2 Optimism intervention in clinical settings ........................................... 204

7.5.3 Optimism interventions in children and adolescents ........................... 206

7.6 Research questions ................................................................................... 207

Chapter 8: Optimism interventions for depression in first-year college students

.................................................................................................................................. 211

8.1 Study 1: individual optimism interventions with depression ................... 211

8.1.1 Intervention designs ............................................................................. 211

8.1.2 Method ................................................................................................. 213

8.1.3 Results .................................................................................................. 216

8.1.4 Discussion ............................................................................................ 223

8.2 Study 2: group optimism interventions with depression .......................... 225

8.2.1 Intervention designs ............................................................................. 225

8.2.2 Method ................................................................................................. 226

8.2.3 Results and analysis ............................................................................. 228

8.2.4 Discussion ............................................................................................ 235

8.3 General discussion ................................................................................... 235

Chapter 9: Understanding optimism .................................................................... 237

9.1 Summary of main findings ....................................................................... 238

9.2 Does culture make a difference ................................................................ 243

9.3 Do people exhibit bias in attributing causes to events happening to others?

246

9.4 Effective optimism interventions for depression ..................................... 248

9.5 Deeper understanding of optimism: theoretical contributions to optimism

literature and future directions ............................................................................. 250

9.6 Is optimism always good? Is pessimism always bad? The evolutionary

explanations for optimism and pessimism ........................................................... 253

Reference ................................................................................................................. 258

6

Acknowledgements

This thesis took a long time, and I have accumulated a debt of gratitude to many

people. I am grateful beyond measure to my supervisors, Timothy Bates and

Alexander Weiss for their support – a constant source of inspiration and

encouragement.

I would like to thank Timothy Bates and Tom Booth for helpful advice on modelling

and Shaoxian Zhou, Jimei Dong, Honejie Tian and Jianjian Teng for their support in

data collection. I am grateful to Dr. Martin Seligman and Dr. Carol Ryff who

authorized me to use their scales in my research.

Thanks to members of the wonderful Differential Club. The weekly meeting made it

possible for me to share updated information and new trends in differential

psychology with them.

I also had amazing support from my family. They provided wonderful emotional

support and good advice. And most of all, thanks to Lily, who shared every moment

in the last three and half years and gave me the confidence to keep going.

7

Declaration

I hereby declare that I am the author of this thesis and that the work presented herein

is my own. This work has not been submitted for any other degree or professional

qualification.

8

Published works

Liu, C., & Bates, T. C. (2014). The structure of attributional style: Cognitive styles

and optimism–pessimism bias in the Attributional Style Questionnaire.

Personality and Individual Differences, 66, 79-85.

9

Abstract

I present seven empirical studies that investigate two main themes regarding two

main approaches of optimism: explanatory style and dispositional optimism. The first

theme incorporates measurement issues and conceptual ideas of optimism and the

second involves optimism interventions on depressive symptoms. In Study 1 I

explored the potential psychometric structure of causal attributions and dispositional

optimism. Attributions may be best viewed as reflecting large differences in

cognitive style, and smaller independent positive- and negative-event biases. For

dispositional optimism, a two-factor model was supported. Study 2 examined

correlations between optimism and the Five-Factor Model of personality.

Dispositional optimism and explanatory style had similar association patterns with

personality, although there were some differences. Study 3 tested and supported a

model in which dispositional optimism mediates the link between explanatory style

and psychological well-being. Study 4 compared the levels of optimism expression

in two ethnic groups, finding that Mainland Chinese participants were more

optimistic and less pessimistic than White British. Study 5 examined attributional

biases and found that individuals show more optimistic biased style for themselves

than for other people. Studies 6 and 7 tested effectiveness of optimism interventions

on depressive symptoms. It demonstrated that self-monitored optimism interventions

on a daily basis could effectively reduce depressive symptoms and increase

optimistic explanatory style. Taken together, the studies replicated some previous

investigations regarding measurement issues and conceptual ideas of optimism, and

explored novel approaches to examining the essence of attributional bias and

effectiveness of optimism interventions in depression treatment. My investigation of

attributional bias is the first to test this idea using new and comparable measures of

attributions. Practicing self-administered optimism interventions is, to my knowledge,

also the first time these interventions have been applied in a sample with mild-to-

moderate depressive symptoms. This may provide an easily monitored and low-cost

alternative to traditional treatments of depression.

Understanding Optimism

Chapter 1: What is optimism 1

Chapter 1: What is optimism?

The optimist sees the rose and not its thorns; the pessimist stares at the thorns,

oblivious to the rose. – Kahlil Gibran (1951, p. 45)

1.1 Origins and concepts of optimism

Optimism from a philosophically historical view

As originally forwarded by Aristotle and as long noted by philosophers afterwards,

human beings are not merely what they are (actuality), but more essentially are what

they are not yet but can be (potentiality) (Chang, 2001a). This idea has been

prominently reflected in the subsequent literature of important philosophers. It was

believed that it is the power of potentiality that determines who and what we are and

how we exist in the world. Here the potentiality means that the range of possibilities

between the two opposite expectations of good or bad things happening, are

outstanding.

Though the roots of psychological accounts of optimism are believed to have

originated from the attempts of leading philosophers of the modern period (Domino

& Conway, 2001), the development of philosophical understanding of optimism can

be traced back to the articulations of the French philosopher Descartes (1596-1650),

who claimed “there is no soul so weak that it cannot, if well-directed, acquire an

absolute power over its passions” (Descartes, 1985).

The original sense of optimism comes from the Latin word optimum,

meaning ‘the best possible’, and technically has its roots in the writings of Gottfried

Leibniz (1646-1716). Leibniz (2010) believed it was God who created the universe

and described it as “the best of all possible worlds.” The term optimism was used to

name the unique maximum or minimum instance of an infinite class of possibilities

in his description. Later, several famous philosophers, including David Hume (1711-



1776), Georg Wilhelm Friedrich Hegel (1770-1831), and Friedrich Nietzsche (1844-

1900), all contributed to the development of psychological accounts of optimism

(Domino & Conway, 2001).

Psychologists have begun to pay attention to optimism from a philosophical

perspective as well. Though Sigmund Freud (1856-1939) was best known for his

pioneering and fundamental work in psychoanalysis, in later life he dedicated his

career to communicating a better social and anthropological understanding of his

essential psychoanalysis principles, which included the philosophical and

psychological status of optimism and pessimism. Freud (1961) claimed that striving

for happiness is in the nature of humans. This process is completed in two

simultaneous-existing forms: an individual wishes to feel extreme joy in life

experience and to avoid distress at the same time. Influenced by the political success

of Hitler’s Nazi party in the 1930s in Germany, Freud shifted from his originally

sceptical view for the future to being deeply pessimistic about the future of humans

(Domino & Conway, 2001). Another pioneering psychologist, William James (1842-

1910), felt similarly pessimistic towards the happiness of humans. However, James

put more emphasis on the individual level, claiming that only each individual has the

ultimate choice between optimism and pessimism (James, 1985).

The philosophical explanation of the origins and development of optimism

are still in progress. All the ideas illuminated above have contributed to our current

understanding of the nature of optimism theoretically. Many theorists have discussed

optimism in human nature in positive terms. One of the useful definitions of

optimism was contributed by anthropologist Tiger (1979, p. 18): “a mood or attitude

associated with an expectation about the social or material future – one which the

evaluator regards as socially desirable, to his [or her] advantage, or for his [or her]

pleasure”. Partly based on this definition, Peterson (2000a) regarded optimism as a

three-factor construct with cognitive, emotional and motivational aspects.

As stated above, optimism has long been discussed in positive terms as

generalized human nature by philosophers and theorists like Descartes, Leibniz,

Hume, and Hegel (Domino & Conway, 2001). At the same time, differential



psychologists began to address optimism as an individual difference, a trait people

possess to varying degrees. Though these two approaches of optimism, human nature

and individual difference, are basically consistent; the differential perspective

focuses more on the influence of an individual’s experience to the characteristic

optimism. Treating optimism as an individual difference means that it is a person’s

experience that influences whether one is optimistic or pessimistic (Peterson, 2000a).

Dictionary definitions of optimism

The Oxford Dictionary provides two related definitions of optimism. The first is

“hopefulness and confidence about the future or the success of something”. The

second conception seems a little bit broader, referring to the belief that “this world is

the best of all possible worlds”. Along the lines of the first definition, Scheier and

Carver (1987) identified optimism as dispositional optimism. Dispositional optimism

refers to positive expectations in a given situation (Scheier & Carver, 1987) and

recently has been conceptualized as broad and general expectancies (Scheier &

Carver, 1992, 1993). Following the second definition, the term optimism has been

applied to the habitual way that people explain their life events, and was identified as

an explanatory style (Seligman, 1991) or attributional style (Abramson, Seligman, &

Teasdale, 1978; Peterson et al., 1982).

While many other competitive models of optimism have been proposed, such

as the Hope construct (Snyder, 1989, 2002; Snyder et al., 1991), the leading

approaches of optimism are explanatory style and dispositional optimism (Carver,

Scheier, & Segerstrom, 2010; Forgeard & Seligman, 2012). These two concepts and

theoretical themes are my main concerns in this research. I will now turn to explicitly

describe these two models.



1.2 Explanatory Style

It has been claimed that individuals are naive psychologists who try to explain the

causes of their own behaviours and those of others (Heider, 1958). One of the

prevailing ideas in psychology is, then, that individuals inherently tend to come up

with explanations for behaviours and outcomes in their lives (Peterson, 2000a).

These views form the foundation of attributional theory. Attributions are taken as the

thoughts and beliefs people hold about the relationships between various

observations and life events, especially those thoughts and beliefs that seek to

explain causal relationships (Poropat, 2002).

1.2.1 Historical Development of models of explanatory Style

The development of attributional theory has a considerable history.

Three dimensions of attributional style

Research on attributional style is widely considered (Abramson et al., 1978) to have

begun with Heider (1958). Heider differentiated internality and externality as

perceived determinants of outcomes. Internality involves explanations “within the

person”, which occur when an individual blames him- or herself for a problem. By

contrast, external explanations turn for causal influences to factors “within the

environment”. These are exemplified in cases when one blames something outside of

oneself.

The next major enlargement of theories of explanatory style came with

Weiner (1974), who added stability as a second attributional component. According

to Weiner, stability refers to attributions about the consistent causes, for instance,

whether the cause is enduring or fleeting. The final enlargement, forming

attributional style theory as it exists today, was initiated by Abramson et al. (1978).

They proposed a three-dimensional model, which incorporated dimensions of

internality-externality, stability-instability, and globality. In this theory, internal and

external attributions resemble the framework of Heider (1958). Stable and instable

attributions are parallel with and the theory of Weiner (1974). Globality, the novel

attributional factor in this theory, is linked to predictions about how likely a causal



factor is to operate across a broad range of additional situations. These three

dimensions, internal versus external, stable versus unstable, and global versus

specific, have been combined to form the three-dimensional model of explanatory

style (Abramson et al., 1978).

During the 1980s, attributional style became a widely-accepted way of

defining and measuring optimism as an individual difference, and much of the

current research on attributions has been inspired by work on this three-dimensional

model of attributional style.

Development of the theory of explanatory style

During the early studies of Maier and Seligman (1976) with animals, it was found

that after being exposed to uncontrollable aversive stressors, animals give up and

become helpless, and later continue to act helpless even when the uncontrollable

negative situations are now under control. This similar phenomenon was tested and

supported on humans as well in later studies (Hiroto & Seligman, 1975; Klein,

Fencil-Morse, & Seligman, 1976), and was called “learned helplessness”. It was

presumed that after experiencing uncontrollable negative events, animals and people

become helpless because they have “learned” that there is no difference in responses

and their subsequent consequences (Maier & Seligman, 1976). Furthermore, this

learning is developed into a generalized expectation that it is futile to attempt a

different future by any action. Helplessness then occurs later following this

pessimistic generalized expectancy of action-outcome independence.

It has been found that certain individuals respond pessimistically after being

exposed to uncontrollable aversive events, while other individuals never give up and

become helpless in similar situations. To account for the different responses of

human helplessness following uncontrollable adversities, the three-dimensional

model of explanatory style was added to the original learned helplessness model.

(Abramson et al., 1978; Peterson et al., 1982). Theory of explanatory style assumes

that causal explanations for a negative event definitively determine whether a person

will develop general helplessness or not. If an individual attributes adversity to an



internal cause, self-esteem is thought to suffer. If they attribute adversity to long-

lasting (stable) causes, helplessness is thought to be enduring. If they attribute a

negative event to a global cause, helplessness is regarded as pervasive (Abramson et

al., 1978; Peterson et al., 1982).

Based on ideas of explanatory style, the reformulated learned helplessness

theory (Abramson et al., 1978) was developed. According to this theory, people

usually search for an explanation for events, especially negative ones occurring in

their lives. Explanation for negative events can vary along three dimensions: internal

versus external, stable versus unstable, and global versus specific (Abramson et al.,

1978). Later on, Seligman (1991) developed research of learned helplessness into

learned optimism by reframing the theory of explanatory style. Thoughts of

helplessness were transformed into optimistic explanatory style, or simply optimism.

Individuals may view negative events as having causes which are unstable, specific,

and external (an “optimistic explanatory style”) or as stable, global, and internal – a

pessimistic explanatory style (Buchanan & Seligman, 1995; C. Peterson & Steen,

2009). People who hold an pessimistic explanatory style will feel pessimistic and be

more prone to depression as a consequence (Peterson & Seligman, 1984). By

contrast, An individual who is characterized with an optimistic explanatory style

appears to be protective for depression (Seligman, 1991).

Generally speaking, explanatory style refers to habitual explanations people

provide for the causes of positive and negative events in terms of their stability,

globality, and internality (Peterson et al., 1982). As these explanations are predicted

to influence behaviour and mood – in particular depression – they are of clinical as

well as theoretical importance (Buchanan & Seligman, 1995; C. Peterson & Steen,

2009).



1.2.2 Measures of explanatory style

Explanatory style or attributional style is mainly reflected in the Attributional Style

Questionnaire (the ASQ; Peterson et al., 1982), which is the associated self-report

measure of attributional style. As Peterson et al. (1982, p. 288) said, ASQ ‘yields

scores for individual differences in the tendencies to attribute the causes of bad and

good events to internal (versus external), stable (versus unstable), and global (versus

specific) factors.’ Accordingly, this self-report questionnaire was developed to assess

the habitual explanation of life events in terms of the stability, globality, and

internality of the causes of positive and negative events (Peterson et al., 1982; 2011).

This questionnaire includes six positive events (e.g., “You do a project that is

highly appraised”) and six negative events (e.g., “You have been looking for a job

unsuccessfully for some time”). Each of these 12 different hypothetical events is

followed by a series of 4 questions which are arranged in the same order.

Respondents are asked to generate an explanation for each event (the first question),

and then to rate this explanation along three dimensions (the remaining three

questions): internal versus external, stable versus unstable, and global versus specific.

These three dimensions, internality, stability, and globality, are defined respectively

as “factors within the person or within the environment” (Heider, 1958), “the degree

of temporal consistency of the cause” (Scheier & Carver, 1987), and “the extent to

which the cause is perceived to recur in other situations” (Abramson et al., 1978).

Basically, the ASQ yields composite scores for explanatory style for positive

events (CoPos, CP, or ASQ Positive) and negative events (CoNeg, CN, or ASQ

Negative); as well as scores for six subscales (Internal Positive, Stable Positive and

Global Positive; Internal Negative, Stable Negative, and Global Negative). To

calculate an overall composite score (CPCN or ASQ Total) of explanatory style, the

negative-event composite is subtracted from the positive-event composite.

Based on responses to these three dimensions for each ASQ event, the subject

is assigned an optimistic or a pessimistic explanatory style. An optimistic

explanatory style consists of explaining positive events as enduring, global and

internally generated, while also explaining negative events as unstable, specific, and



externally caused (Forgeard & Seligman, 2012). Reflected in the measuring and

scoring of the ASQ, a positive score of CPCN represents an optimistic explanatory

style and a negative score of CPCN represents a pessimistic explanatory style.

Optimistic explanatory style scores have been linked to protection from depression

(Peterson & Seligman, 1984) and physical illness (Wise & Rosqvist, 2006) as well as

higher academic achievement, subjective and physical well-being, and career

achievement (Forgeard & Seligman, 2012).

Psychometric properties of the ASQ

Within attributional models of depression, the attributions are seen as causing

distinct behavioural consequences. For example, low self-esteem is predicted to

result from internal attributions regarding negative events, while chronic depression

is suggested to result from stable attributions for negative events (Peterson et al.,

1982). In this model of learned helplessness, depression emerges as a consequence of

experience with uncontrollable negative events (Abramson et al., 1978). The concept

of attributional style, however, predicts that the three types of explanation (internality,

stability, and globality) are correlated with each other within at least each event

valence.

However, more recent research based on this model has resulted in findings that

are somewhat counterintuitive. One of the earliest studies dealing with this question

was conducted by Peterson et al. (1982). They reported that attributions for positive

events and attributions for negative events were uncorrelated (r = .02). This lack of

correlation between explanatory styles for positive and negative events has been

found in other work as well. For example, P.J. Corr and J.A. Gray (1996) examined

the factor structure of the ASQ in two independent samples using Varimax rotated

principal components analysis. They found that positive and negative explanatory

styles were independent. Additionally, whereas for negative events, internality

ratings were largely independent of stability and globality ratings, for positive events

these three dimensions formed a single factor, suggesting that explanations for

positive and negative events might have different structures.



Succeeding studies have used larger samples, and incorporated confirmatory

structural equation modelling (SEM), allowing a better understanding of the structure

of attributions by contrasting competing theoretical models. For instance, Higgins,

Zumbo, and Hay (1999) reported a confirmatory factor analysis of the ASQ

identifying three correlated factors in a sample consisting of more than 1,000

subjects. This model was a good fit for attributions of both negative events and

positive events. Consistent with several other studies, the stability and globality

factors correlated strongly, with internality-externality being more independent of the

globality in this study.

Multi-method analytic strategies were incorporated later in attributional style

SEM analysis since it was realized that subjects are generating multiple responses to

each ASQ event. This is an important innovation, as misleading results can arise in

analyses of data generated from multiple correlated responses based on each item,

and it is true in the ASQ where all three attributions are samples for each event.

Based on this multi-method analysis strategy, it was confirmed that the three-

dimension structure of explanatory style still provided a good account of responses to

negative events in terms of correlated latent factors of internality-externality,

stability-instability, and globality-locality (Hewitt, Foxcroft, & MacDonald, 2004).

However, this model indicated higher correlations between internality and the other

two factors for negative events.

Other measures of explanatory style

In addition to the most widely-used tool, the ASQ, several other measures have been

developed to assess explanatory style. Most of these measurements are designed on

the basis of similar criteria and scoring method with the ASQ, though they consist of

different events or are adapted to suit subjects with diverse backgrounds. The

Expanded Attributional Style Questionnaire (EASQ;Peterson & Villanova, 1988) is

one such tool. The EASQ yields the same composite and subscale scores as the ASQ,

but contains only 24 negative events, each of which subjects indicate a cause of the

event and rate the three dimensions of internality, stability, and globality of the cause

on 7- point Likert scales. The EASQ is claimed to be a better measure in



investigations of the reformulated learned helplessness theory than the ASQ, since it

is believed that people’s explanatory style for negative events connects highly with

helplessness and depression (Metalsky, Joiner, Hardin, & Abramson, 1993) .

Based on the reformulated helplessness theory of depression (Abramson et al.,

1978), Abramson and Metalsky (1989) developed the self-report Cognitive Style

Questionnaire (CSQ) as another modified and expanded version of the ASQ. The

CSQ made two modifications to the ASQ. First, ratings of the probable consequences

and self-worth implications were added to each hypothetical event, which make it

possible to measure all three components of the cognitive vulnerability factor implied

in the reformulated learned helplessness theory. Second, the hypothetical events were

extended to include 12 positive and 12 negative events in the CSQ. In a review with

30 studies, Haeffel et al. (2008) reported the psychometric and validity properties of

the CSQ.

In addition to generally widely-accepted measures of explanatory style listed

above, there are some other explanatory style measures developed in specific

domains of different backgrounds (for a review, see Smith, Caputi, & Crittenden,

2013), such as the Academic Attributional Style Questionnaire (AASQ; Peterson &

Barrett, 1987), the Sport Attributional Style Scale (SASS; Hanrahan, Grove, & Hattie,

1989), the Team Attributional Style Questionnaire (TASQ; Shapcott & Carron,

2010), and the Workplace Explanations Survey (WES; Smith et al., 2013). The most

widely used measure for assessing children’s explanatory style is the Children’s

Attributional Style Questionnaire (CASQ; Kaslow, Tannenbau, & Seligman, 1978).

The CASQ consists of 24 positive and 24 negative hypothetical events. This

instrument has the same construction and format as the original ASQ.

The ASQ, the EASQ, the CSQ, the AASQ, the SASS, the TASQ, the WES,

and the CASQ are all self-report measures, among which the ASQ has been most-

widely used in application. The second popular way of assessing explanatory style is

the Content Analysis of Verbatim Explanations (CAVE; Peterson, Berres, &

Seligman, 1985) technique. This instrument was developed to assess explanatory

style by analysing statements, journal entries, speeches, and other written materials



which are believed to contain causal explanations. The CAVE has been frequently

used in studies of explanatory style and physical well-being considering its

advantage in longitudinal research (Peterson, 1988).

1.2.3 Stability and heritability of explanatory style

Is explanatory style a relatively stable personality trait? Are attributions stable

enough across time and situations to guarantee the existence of the designated

explanatory style? To answer these questions, the consistency of explanatory style

has been explored by several studies, which suggest that there is at least some

stability in attributional style over time and circumstances. For example, in a study

conducted by Tiggemann, Winefield, Winefield, and Goldney (1991), explanatory

style was measured in young adults across a period of three years. The results

showed that explanatory style tested in the first time period was moderately

correlated to explanatory style measured in the second (r = .44).

In another longitudinal study, Burns and Seligman (1989) reported that

explanatory style for negative events during early adulthood was positively related to

explanatory style for negative events 52 years later (r = .54), and the dimension of

stability accounted for most of the observed correlations of explanatory style for

negative events. Explanatory style for positive events, however, was not as stable as

that for negative events. The composite positive score at baseline was not

significantly correlated with the same test at 52 years later (r = .13).

The stability of explanatory style can be partly explained by its heritability or

the influence of biological factors on this trait. So far as I know, not many genetic

studies have been done to explore the heritability of explanatory style. In one

exception, Schulman, Keith, and Seligman (1993) conducted a pioneering twin study

with a sample of 115 pairs of identical twins and 27 pairs of dizygotic twins.

Participants were directed to complete the ASQ. The composite score summing up

responses to both positive and negative events (CPCN), the scores for the sub-scale

of negative events (CN), and reactions to the positive events (CP) were analysed

separately. For CPCN, the correlations were .48 for identical twins and 0.00 for

dizygotic twins, which suggests a substantial hereditary effect of explanatory style.



For CN, the correlations were .43 for identical twins and -.03 for dizygotic twins,

showing the same pattern as CPCN. In contrast, the scale for positive events (CP)

also showed a moderate correlation of .50 for identical twins. Comparatively,

however, the correlation for dizygotic twins was nearly as high (.41), which might

demonstrate a substantial effect of shared environment. The different patterns

suggest that heritability of explanatory style may be indirect.

1.2.4 Self-serving attributional bias and optimistic explanatory style

People have a need to view themselves positively. This is easily the most common

and consensually endorsed assumption in research on the self. – Heine, Lehman,

Markus, and Kitayama (1999, p. 766).

As one of the most important psychosocial systems of optimism, explanatory style or

attributional style has been the subject of a large body of research, which provides

consistent evidence for the linkage between this trait and many other psychological

traits. Such attributions can be functional and adaptive and may serve psychological

and social purposes when attributional bias applies (Mezulis, Abramson, Hyde, &

Hankin, 2004; Sanjuan & Magallares, 2014). This comes along with the proposal of

positive cognitive bias of human nature (Heider, 1958) and much prior research

concerning individuals’ biased attributions to happenings in their lives (Cadinu,

Arcuri, & Kodilja, 1993). Though attributional bias and explanatory style basically

share similar measures and scoring methods currently, they have been proposed and

studied mostly separately.

Attributional bias was argued to be manifested in two related and different

modes. One is self-serving attributional bias, which refers to the tendency of

individuals to explain negative events or outcomes with more external or contextual

causes, while attributing positive events or outcomes to more internal or controllable

causes (Mezulis et al., 2004). The other form of attributional bias is self-other bias,



assuming that individuals tend to promote a favourable perception in attribution of

the self in comparison to others (D. T. Miller & Ross, 1975). This tendency of self-

serving attributional bias is pervasive in the general population across age, ethics,

and psychopathology (Mezulis et al., 2004).

The theoretical basis of self-serving bias in attribution derived from the

interaction between motivation and cognition certainty, suggesting that people tend

to “accept responsibility for positive behavioural outcomes and to deny responsibility

for negative behavioural outcomes” (Bradley, 1978, p. 59). Prior studies addressing

self-serving attributional bias are quite varied in the measures and thus in the

operational definitions of this bias. This self-serving bias used to focus on the

attributional dimension of internality by assuming that individuals exhibit more

internal attributions for positive events than for negative events (Greenberg,

Pyszczynski, & Solomon, 1982; Nurmi, 1992).

With the development of the most widely-used measure of attributions, the ASQ,

it has been debated that it is insufficient to establish a self-serving attributional

pattern only using the internality dimension. Accordingly, this self-serving bias has

been extended to also include the other two dimensions of attributions, namely

stability and globality. Self-serving attributional bias is consequently conceptualized

as the tendency of people to attribute positive situations to more internal, stable, and

global causes than for negative situations (Mezulis et al., 2004).

Though self-serving attributional bias and optimistic explanatory style have been

reported separately in most of previous studies, these two concepts have similar

definitions since self-serving bias has been conceptualized within the three-

dimensional model of attributions. While an optimistic explanatory style consists of

explaining positive events as enduring, global and internally generated, while also

explaining negative events as unstable, specific, and externally caused (Forgeard &

Seligman, 2012); self-serving attributional bias is defined as the tendency of people

to attribute positive situations to more internal, stable, and global causes than for

negative situations (Mezulis et al., 2004). Accordingly, an optimistic explanatory

style is a positive pattern consistent with self-serving attributional bias defined above,



or, in other words, self-serving attributional bias is the universal positive bias in

explanatory style.

Evidence of interchangeability between these two concepts is found in the

similarity of measuring and scoring as well. Basically, the ASQ and the adaptation

versions of the ASQ were among the most commonly used self-report measures in

prior studies of self-serving attributional bias (for review, see Mezulis et al., 2004).

While a more “optimistic” attributional style for a domain means higher scores for

positive events and a lower score for negative events for that domain (Forgeard &

Seligman, 2012), a self-serving attributional bias represents a positive score when

attributions for negative outcomes are subtracted from attributions for positive

outcomes (Sanjuan & Magallares, 2014). Specifically, on one hand, if the subtraction

score of the ASQ Negative from the ASQ Positive is positive, it represents a self-

serving attributional bias or an optimistic explanatory style, reflecting stronger

attributions along internal, stable and global causes for positive than for negative

events. On the other hand, if the subtraction score of the ASQ Negative from the

ASQ Positive is negative, it then stands for lack of a self-serving attributional bias or

an optimistic explanatory style, reflecting weaker attributions for positive than for

negative events (Sanjuan & Magallares, 2014).

Moreover, prior research along both lines of optimistic explanatory style and

self-serving attributional bias is consistent in their findings of beneficial influences

on well-being (Forgeard & Seligman, 2012; Mezulis et al., 2004). For reasons of

consistency, in my research of positive bias in attributions, the tendency of holding

an optimistic explanatory style and the tendency of expressing a self-serving

attributional bias will be referred to as equal to each other, both referring to the

tendency of individuals to explain positive situations through internal, stable and

global causes, and negative situations to external, unstable and specific causes. That

is, self-serving attributional bias is taken as the tendency of holding an optimistic

explanatory style in explanation of positive and negative events normally specified in

the ASQ.



1.2.5 Explanatory style, hopelessness, and depression

Hopelessness is an important concept in establishment and development of the

hopelessness theory of depression (Abramson, Metalsky, & Alloy, 1989), in which

depression is conceptualized as an overabundance of negative moods and negative

cognition. According to the hopelessness theory of depression, hopelessness is

conceptualized as the expectancy that future outcomes will be stable, global, and will

negatively influence many aspects of an individual’s life regardless of his or her

efforts (Abramson et al., 1989). As a result, hopelessness about the future constitutes

a sufficient and proximal cause of a subtype of depression, called hopelessness

depression (Abramson et al., 1989). ‘The hopelessness theory represents a theory-

based approach to the classification of a subset of the depressive disorders and

postulates the existence in nature of hopelessness depression…’ (Abramson et al.,

1989, p. 359).

Abramson et al. (1989) pointed out that hopeless depression are more likely

to occur when negative events are attributed to stable and global causes.

Comparatively, the influence of internality dimension is deemphasized when

symptoms of hopelessness depression are discussed. Separation between the

internality dimension and the other two attributional dimensions (stability and

globality) was supported by empirical studies. For instance, Higgins et al. (1999)

reported a confirmatory factor analysis of the ASQ identifying three-correlated

factors in over 1,000 subjects. It indicated that the stability and globality factors

correlated strongly (r = .61 for negative events, r = .67 for positive events), with

internality-externality being more independent of the globality (r = .35 for negative

events, r = .28 for positive events). Thus, in ASQ, Hopelessness (stability + globality

of negative events) is produced as a composite score separately from other composite

scores.

This attributional model of depression has accumulated substantial evidence

from empirical studies (e.g. Vazquez, Jimenez, Saura, & Avia, 2001). For instance,

295 secondary school students were instructed to complete measures of attributional

style, self-esteem, and depression (Kurtovic, 2012). This study indicated that

hopelessness correlated significantly with depression (r = .58). Similarly, Ahrens and

Haaga (1993) reported that hopelessness is significantly correlated with depressive



symptoms (r = .20) (Peterson & Vaidya, 2001). Cross-sectional studies propose that a

pessimistic attributional style is correlated with hopelessness and thus depression. On

the other hand, an optimistic explanatory style has been linked to protection from

depression. A pessimistic explanatory style predicts increases in depression over time

in different populations, such as lower-class women, children, and depressed patients

(Peterson & Seligman, 1984). Peterson and Vaidya (2001) reported that hopelessness

positively correlated with depression in their study with a group of college students (r

= .20).

1.3 Dispositional Optimism

As mentioned in the beginning of this chapter, one of the two related concepts of

optimism provided theoretically by the Oxford Dictionary is “hopefulness and

confidence about the future or the success of something”. Consistent with this

dictionary definition and following traditional folk wisdom about optimism, Scheier

and Carver (1987) have studied a personality trait identified as dispositional

optimism. Based on theoretical studies on the expectancy-value model and self-

regulatory model, dispositional optimism originally referred to positive expectations

in a given situation and later was conceptualized as broad and general expectancies

(Scheier & Carver, 1992, 1993).

Framed within the definition of dispositional optimism, being optimistic

means simply that people expect good things to happen to them in the future, and

being pessimistic means that people expect bad experience in the future (Carver et al.,

2010; Scheier & Carver, 1987). It has long been believed that the level of generalized

favourable expectancies for the future is related prospectively with many, perhaps all,

facets of life (Carver & Scheier, 2014). This belief has been supported by a good deal

of systematic studies in the past 25 years or so (Carver et al., 2010; Scheier & Carver,

1987, 1992).

1.3.1 Historical development of models of dispositional optimism



The perspective of “dispositional optimism” originated theoretically from the

expectancy-value model and has been developed from research conducted by Scheier

and Carver (2001). There is a long history of theoretical research on motivation of

behaviour, and two facets have been identified in the proposed expectancy-value

model. On one hand, it is assumed that people act around the pursuit of goals (Austin

& Vancouver, 1996). Goals are states or actions that people take as desirable or

undesirable. The more important a given goal is to an individual, the greater is the

element of value in the person’s motivation to pursue this goal. People have no

motivation to act without having a goal that is valued to some extent. That is, people

are inclined to fit their actions to values they regard as desirable.

On the other hand, expectancy was proposed to be the other conceptual

element of the motivation model (Carver & Scheier, 2001). The assumption of

expectancy is linked to a sense of confidence and doubt about a given goal’s

attainability or avoidability. A person has no desire to take action if he or she lacks

confidence. Only if people have adequate confidence will they move into effort.

Confidence and doubt are also important for a person to continue or quit an action.

Based on this model of motivation, dispositional optimism was proposed and

is seen as a broad and generalized version of confidence and persistence in pursuit of

desirable goals (Scheier & Carver, 1992). It is assumed that optimism should be

continuous even when progress is difficult or slow in the face of adversity (Carver et

al., 2010). According to Carver and Scheier (2001), virtually all fields of human

activity can be defined in term of goal pursuit, and people’s thoughts and actions

imply the identification and adoption of goals and the adjustment of behaviour

toward these goals. As a result, Carver and Scheier (2001) refer to their perspective

in dispositional optimism as a self-regulatory approach. To be specific, optimism

enters into a self-regulatory loop when people ask themselves about the obstacles to

pursuing the goals they have adopted. Do people still believe they can achieve their

desirable goals in the face of impediments? Optimists and pessimists are

differentiated depending on their belief. If people are confident in achieving the goals

even in face of difficulties, they are seen as being optimistic; if not, they are

pessimistic individuals.



Carver et al. (Carver et al., 2010; Scheier & Carver, 1992) stated that

“optimism and pessimism are confidence and doubt […] pertaining to life, rather

than to just a specific context”. Here we can see that optimism and pessimism are

broad, generalized versions of expectations to future life, rather than to just a specific

narrow context. And this generalized confidence or doubt will continue during actual

behaviour even in the face of difficulties.



1.3.2 Measures of dispositional optimism

To assess dispositional optimism, researchers ask people directly whether they

expect outcomes in their future lives to be beneficial or unbeneficial (Scheier &

Carver, 1992). This way of assessing dispositional optimism is acquired by using

self-report questionnaires such as the Life Orientation Test (LOT; Scheier & Carver,

1985) or its successor the Life Orientation Test-Revised (LOT - R; Scheier, Carver,

& Bridges, 1994).

The LOT consists of 12 items (four filler items included), in which four are

described in a positive direction (e.g., “I always look on the bright side of things”),

and four in a negative direction (e.g., “I rarely count on good things happening to

me”). Respondents are directed to assess the extent to which they agree with each of

the 12 items on a 5-point scale (4 = strongly agree, 3 = agree, 2 - neutral, 1 =

disagree, and 0 = strongly disagree).

The LOT was revised later to resolve indistinguishable problems among

dispositional optimism and other personality traits, such as Neuroticism (Scheier et al.

(1994). Two originally problematic (positively worded) items were eliminated. To

keep the scoring balance between positively worded and negatively worded items,

one new positively worded item was added and one negatively worded item was

removed. As a result, the LOT-R consists of 10 items (four filler items included), in

which three items are keyed in a positive perspective and three in a negative

direction. For each item, respondents assess their levels of agreement or

disagreement on a 5-point scale.

Scheier and Carver (1985) originally proposed the LOT to measure a one-

dimensional bipolar construct of dispositional optimism. For LOT-R, (Scheier et al.,

1994) also proposed that “confirmatory factor analysis further indicated that the

single-factor solution was superior to a two-factor one.” However, evidence

indicated that the two-factor model, which declared that optimism and pessimism

represent two distinct traits, was proposed and replicated in many studies later

(Chang, Maydeu-Olivares, & D'Zurilla, 1997; L. Chang & McBrideChang, 1996;

Creed, Patton, & Bartrum, 2002; Roysamb & Strype, 2002). The applicability of this



two-factor model was also supported in studies with Eastern subjects (Cheng &

Hamid, 1997; Li, 2012; Sumi, 2004).

1.3.3 Stability and heritability of dispositional optimism

Stability of dispositional optimism

Is dispositional optimism a relatively stable personality trait across time and

situations? How consistent is an individual’s level of dispositional optimism? As

with most personality traits, test-retest reliabilities are relatively high in several

longitudinal studies (although it is not always the case). For instance, within a group

of 182 middle-generation women, Atienza, Stephens, and Townsend (2004) found

the LOT test-retest correlation of .73 across a one-year period.

Lucas, Diener, and Suh (1996) conducted one study across a period of four

weeks, during which 212 college students were required to assess their dispositional

optimism twice using the LOT. The test-retest correlation of dispositional optimism

between the two periods was .76. Also, with a group of 82 college students,

Billingsley, Waehler, and Hardin (1993) reported a test-retest correlation of .78 for

the LOT across a period of four weeks. In the pioneering study of LOT formation,

Scheier and Carver (1985) found an even higher test-retest correlation of .79., based

on assessments of 142 participants within a four-week interval. Studies conducted in

Eastern societies have also reported the stability of the LOT and the LOT-R. For

instance, in a Hong Kong Chinese sample, test-retest reliability coefficients across a

period of five months were reported as .68 for the LOT and .66 for the LOT-R (Lai,

Cheung, Lee, & Yu, 1998).

However, research results on consistency of dispositional optimism over

longer time periods are controversial. For example, in a study across a 10.4 year

period in a group of 209 middle-aged women, Matthews, Räikkönen, Sutton-Tyrrell,

and Kuller (2004) found a test-retest correlation of .71, similarly to other traits.

However, in another 10-year-period study conducted by Suzanne C. Segerstrom

(2007), the LOT test-retest correlation of dispositional optimism was only .35.



Though there were less than 100 participants, the result nevertheless indicated that

change in dispositional optimism is possible at least for some people.

Heritability of dispositional optimism

The definition of dispositional optimism as a general tendency to have positive or

negative expectancies (Scheier & Carver, 1987) is compatible with ideas of

evolutionary psychology addressing the general characteristics of a species.

To test the heritability of dispositional optimism, Plomin et al. (1992)

conducted the pioneering study in a sample of more than 500 same-sex pairs of

middle-aged identical and fraternal twins, half of whom were reared together (126

pairs of identical and 146 pairs of fraternal twins) and half raised apart (72 pairs of

identical and 178 pairs of fraternal twins). Participants took the LOT twice over a

period of three years. For identical twins reared apart, the correlations indicated

heritabilities of 23% for LOT optimism and 27% for LOT pessimism. As expected,

the correlations for identical twins raised together were lower, 39% for LOT

optimism and 20% for LOT pessimism. Generally speaking, a heritability of 25% for

optimism was reported in this study. Similarly, in a sample consisted of 428 Italian

twin pairs (aged 23-24 years), Caprara et al. (2009) reported a heritability of 28% for

dispositional optimism.

Research on the heritability power of dispositional optimism has also

conducted in much larger samples. For instance, Mosing, Zietsch, Shekar, Wright,

and Martin (2009) measured dispositional optimism in 3,053 Australian twins

(ranging in age from 50 to 94 years) using the LOT over 50 years. The sample

included 501 identical female twins, 153 identical male twins, 274 dizygotic female

twins, 77 dizygotic male twins, 242 dizygotic opposite-sex twin pairs, and 561 single

twins (without participation of the co-twin). This study revealed that additive genetic

factors explained 36% of the variation in dispositional optimism. This sample was

combined with 406 pairs of Swedish twins later to analyse the relationship between

dispositional optimism and mental health (Mosing, Pedersen, Martin, & Wright,

2010). A heritability estimate of 34% for dispositional optimism was reported in this



combined sample. Another twin study conducted by Mosing et al. (2012) explored

the relationship between dispositional optimism and longevity, and it indicated that

the association between dispositional optimism and longevity may have genetic

involvement as well.

In addition to genetic behavioural studies that directly investigate the

heritability power of dispositional optimism, some other studies offered evidence

using different approaches. For example, in a study with two separate population-

based cohorts, Rius-Ottenheim et al. (2012) reported that parental longevity was

positively associated with dispositional optimism in adult offspring, indicating some

sort of genetic underpinning in this personality trait. Later, J. J. Yu and Ko (2013)

investigated the link between father’s and child’s dispositional optimism in a sample

of 422 father-child dyads in South Korea. It found that father’s dispositional

optimism was positively correlated with child’s dispositional optimism (r = .55).

These kinds of studies support the heredity of dispositional optimism from the aspect

of generation transmission.

Previous research based on twin studies suggests that heritabilities of

dispositional optimism (.25-.36) are not that high (compared with typical personality

traits) but statistically significant, indicating that there is stability in dispositional

optimism and an influence of genetic factors on this trait. Attempts to identify

specific genomic elements underlying variations of optimism have shown mixed

results (see review of Carver & Scheier, 2014). It also should be kept in mind that,

like all other personality traits, optimism is still affected by non-shared

environmental effects, or the experiences people have in life.



1.4 Benefits of Optimism

Optimism has had a constantly favourable reputation over the years. A variety of

poets, writers, philosophers, psychologists, and social workers, have described

optimism as greatly beneficial to both individuals and the general world around us

(Chang, 2001b). On the other hand, pessimism is considered as at least contributing

to depression, passivity, morbidity, and failure. It is believed that optimism has had

an adaptive value in dealing with environmental risks and life challenges over the

million or so years of evolution (Tiger, 1979). And, this adaptive advantage of being

optimistic still works for people to achieve more in current life (Seligman, 2011).

Optimism is a cognitive construct intertwined with emotional, motivational,

and behavioural processes, and research of optimism has extended to diverse

directions in psychological studies (Carver & Scheier, 2014). Research over the past

three decades has documented beneficial effects of optimism on enhancing well-

being. A large and growing literature indicates that, no matter how optimism is

conceptualized and measured, it is linked to positive emotions and behaviours; to

prominent physical well-being; to persistence and active coping strategies; to

outstanding academic and occupational performance; and even to resilient and

adaptive social relationships (for reviews, see Andersson, 1996; Carver & Scheier,

2014; Carver et al., 2010; Forgeard & Seligman, 2012; Scheier & Carver, 1992).

Regarding the two optimism models, optimistic explanatory style scores have

been linked to protection from depression (Peterson & Seligman, 1984) and physical

illness (Wise & Rosqvist, 2006) as well as higher academic achievement, subjective

and physical well-being, and career achievement (Forgeard & Seligman, 2012).

Similarly, self-serving attributional bias has also long been positively associated with

mental and physical health (for review, see Mezulis et al., 2004). Not surprisingly,

the studies of dispositional optimism have shown that higher levels of optimism are

correlated with positive life outcomes in various contexts (Carver et al., 2010;

Scheier & Carver, 1987, 1992, 1993). Generally speaking, no matter how optimism

is conceptualized and measured, research is uniform in indicating that optimism is



bonded with beneficial characteristics: happiness, achievement, health, and

persistence.

1.4.1 Optimism and physical well-being

Based on the widely accepted perspective that optimism is generally beneficial in life

of human being, increasing number of physicians has acknowledged the benefits of

thoughts and emotions characterized by optimism on physical well-being (Peterson

& Bossio, 2001; Rasmussen, Scheier, & Greenhouse, 2009).

Explanatory style examines the habitual explanations people provide for

events, and is seen as a distal influence on helplessness and failures of adaption that

involve helplessness (Peterson & Seligman, 1984; Seligman, 1991, 2011). This

expectation of helplessness is theoretically linked to outcomes such as physical well-

being. Empirical studies concerning this issue have been facilitated by development

of widely accepted measures of attributional style, such as the ASQ and CAVE.

Having a higher level of dispositional optimism has also been consistently

involved with better physical health. The potential mechanism underlying this

correlation is that thinking positively about the future may result in a more active

attitude towards the stressors of life than thinking pessimistically, and lower stressor

responses may lead to less physical detriments on the body, and may result in better

physical health as a final result (Carver et al., 2010).

Numerous studies have been conducted to examine the positive link between

optimism (including both explanatory style and dispositional optimism) and physical

health based on both general settings (see reviews by Forgeard & Seligman, 2012;

Kamen & Seligman, 1987; Norvell, 1992; Peterson, 1988, 2000b; Rasmussen et al.,

2009; Scheier & Carver, 1987, 1992; Seligman, 1989; Snyder, 2002) and many

different specific contexts, including the immune system (Suzanne C. Segerstrom &

Sephton, 2010), chronic pain (Goodin & Bulls, 2013), cancer, AIDS (Tomakowsky,

Lumley, Markowitz, & Frank, 2001), cardiovascular health (Bennett & Elliott, 2005;

Giltay, Geleijnse, Zitman, Hoekstra, & Schouten, 2004), carotid atherosclerosis

(Matthews et al., 2004), ambulatory blood pressure (Räikkönen & Matthews, 2008),



coronary heart disease (Tindle et al., 2009), and bone marrow transplantation

(Hochhausen et al., 2007). There is also evidence that optimists show more adaptive

sleep patterns both for children (Lemola et al., 2011) and adults (Lemola, Raikkonen,

Gomez, & Allemand, 2013).

Rasmussen et al. (2009) conducted a meta–analysis using 108 studies

exploring the relationship between optimism (including dispositional optimism and

explanatory style) and physical health, and reported an overall correlation of .18 (p

< .001) between optimism and physical health outcomes, and this correlation

remained significance even after adjusting for Neuroticism and psychosocial factors.

Taken together, optimism is characterized by its health-promoting properties, though

it is still not quite clear what the possible mechanisms are linking optimism and

health.

1.4.2 Optimism and psychological well-being

Well-being has been measured largely in two distinct traditions, hedonic and

eudemonic well-being, or of subjective well-being and psychological well-being,

with the former normally measured with the Satisfaction with Life Scale (SWLS;

Diener, Emmons, Larsen, & Griffin, 1985) and the Positive and Negative Affect

Schedule (PANAS; Watson, Clark, & Tellegen, 1988), and the latter being most

widely implemented using the Ryff scales of psychological well-being (RSPW; Ryff,

1989; Ryff & Keyes, 1995).

While subjective well-being focuses on happiness and pleasure (Diener, Suh,

Lucas, & Smith, 1999), psychological well-being, which stems from the tradition of

eudemonic well-being and was further developed in the field of positive psychology,

emphasizes the fulfilment of human potential (Ryff, 1995). According to Ryff (1989),

psychological well-being is defined by six related dimensions, including autonomy,

environmental mastery, personal growth, positive relations with others, purpose in

life, and self-acceptance.



Explanatory style and subjective well-being

Explanatory style is a distal influence on helplessness and failures of adaption that

involve helplessness. As Wise and Rosqvist (2006, p. 293) said, “Explanatory style

can have a significant and prolonged impact on well-being. Whereas pessimistic

explanatory style can negatively impact several facets of well-being, …, optimistic

explanatory style may serve as a protective factor”.

It seems that individuals with an optimistic explanatory style tend to have

promising expectations for the future, believing that good will prevail and whatever

events are being experienced will all be worthwhile in the end. Moreover, individuals

with an optimistic explanatory style tend to accept stressful experiences because of

this viewpoint. These beliefs and acceptance help individuals who have an optimistic

explanatory style to cope effectively with challenging and demanding situations.

Effective and positive coping then finally facilitates well-being.

The argument that explanatory style predicts well-being arises from many

studies associating depressive symptoms with a pessimistic explanatory style,

measured with the ASQ or the CAVE. For example, Peterson and Seligman (1984)

reviewed a variety of evidence showing that a pessimistic explanatory style predicts

increases in depression over time in different populations, such as lower-class

women, children and depressed patients. Similarly, Ahrens and Haaga (1993)

reported that attributional style for positive events was associated with positive

affectivity (r = .47), and attributional style for negative events was associated with

negative affectivity (r = .21), depression (r = .31), and anxiety (r = .38). In addition,

hopelessness (stability and globality of the ASQ) is significantly correlated with

depressive symptoms (r = .20) (Peterson & Vaidya, 2001).

Additionally, one study conducted on a sample of 280 adults who were

divided into three age groups reported that a pessimistic explanatory style in negative

affiliation domains correlated significantly with depressive symptoms in older adults

(Isaacowitz, 2005). The positive relationship between explanatory style for negative

events and depression was also supported by an SEM-approach study, in which the



correlation between these two variables was reported as .30 (Ledrich & Gana, 2013).

Further, in a recent study, Sanjuan and Magallares (2014) reported the positive

correlations between self-serving attributional bias and higher scores of life

satisfaction (r = .31) and affect balance (r = .46). Longitudinal studies also give

support to the beneficial effect of an optimistic explanatory style on mental health.

For example, in a four-week follow-up study with a group of 167 college students,

Kleiman, Liu, and Riskind (2013) found that an optimistic attributional style

predicted decreased levels of stressful events over the following four weeks, even

when symptoms of depression were controlled for.

The links between explanatory style and subjective well-being have also been

investigated in a wide range of contexts including different stressful situations, such

as heart transplant patients (Jowsey et al., 2012), breast cancer patients (Colby &

Shifren, 2013), and patients with advanced cancer (Applebaum, Stein, Rosenfeld, &

Breitbart, 2012). Results of these studies indicated significant positive association

between an optimistic explanatory style and overall subjective well-being.

Dispositional optimism and subjective well-being

People with a high level of dispositional optimism tend to expect good things to

happen to them in the future, even when confronting difficulties. This general

tendency yields a relatively positive mix of feelings and adaptive coping strategies,

which enhancing subjective well-being and good health (Wrosch & Scheier, 2003).

The relationship between dispositional optimism and subjective well-being

has been investigated in numerous studies, which mainly used the LOT or LOT-R. In

a review of 56 studies (Andersson, 1996), it was reported that the average weighted

correlation between dispositional optimism and depressive symptoms was -.45.

Peterson and Vaidya (2001) also reported that expectations were significantly

correlated with depressive symptoms (r = -.55).

Studies conducted in people in different life stages revealed that being

optimistic is a beneficial property for both young and old people. For example, with

a group of 504 high school students, Creed et al. (2002) found that students with high



level of dispositional optimism scored low on psychological distress. Isaacowitz

(2005) addressed this issue in a wider range with three age groups (young, middle-

aged, and older adults). The study reported that dispositional optimism correlated

with greater life satisfaction and lower levels of depressive symptoms across all three

age groups. It also found that dispositional optimism was correlated with positive

affect (r = .44) in one study with 225 adults aged from 65 to 94 years (Ferguson &

Goodwin, 2010).

Evidence from twin studies provides further support for the positive aspects

of being optimistic. Plomin et al. (1992) reported (n = 500) that dispositional

optimism was significantly associated with depression and life satisfaction (.54 on

average). These associations remain significant even after Neuroticism is controlled

for. This result is further supported by another twin study with a larger sample (n =

1,304). It indicated that dispositional optimism predicts high levels of mental health

(Mosing et al., 2010; Mosing et al., 2009).

Studies conducted on people in stressful situations may better explain the

significant correlations tween optimism and subjective well-being. Those situation-

specific studies involved different groups of participants including gay men with

AIDS (Taylor et al., 1992), skin cancer patients (Luo & Isaacowitz, 2007), patients

with breast cancer (Colby & Shifren, 2013), muscle disease patients (Graham et al.,

2014), freshmen in college (Brissette, Scheier, & Carver, 2002; Chemers, Hu, &

Garcia, 2001; Rand, Martin, & Shea, 2011), ethnic minority adolescents in urban

areas (Vacek, Coyle, & Vera, 2010), women after childbirth (Carver & Gaines, 1987),

and older widows in their first single year (Minton, Hertzog, Barron, French, &

Reiter-Palmon, 2009). In summary, previous research indicates that individuals

scoring high on optimism tests are more likely to perform adaptive, health-promoting

behaviors even when they confronted with challenging situations.

Is this positive correlation between dispositional optimism and subjective

well-being consistent across time? This issue has been addressed in at least one

longitudinal study. This study investigated the effects of optimism on subjective

well-being at two time points over a six-year interval, and reported that being



optimistic was correlated with higher levels of positive affect and life satisfaction

(Daukantaite & Zukauskiene, 2012).

Optimism and psychological well-being

Several studies have reported the positive relationship between dispositional

optimism and psychological well-being. For example, Augusto-Landa, Pulido-

Martos, and Lopez-Zafra (2011) reported in a sample of 217 undergraduates that

dispositional optimism showed significant positive associations with all six

dimensions of psychological well-being (r ranged from .38 to .59).

Similarly, in a study conducted within a group of 225 older adults, Ferguson

and Goodwin (2010) found that dispositional optimism was positively correlated

with Purpose in Life (one of the six dimensions of psychological well-being). The

positive correlation between dispositional optimism and psychological well-being

was reported in an adolescent sample as well. It revealed that LOT-R scores were

positively correlated with all six dimensions of the RPWB (r ranged from .32 to .56)

(Monzani, Steca, & Greco, 2014).

The relationship between explanatory style and psychological well-being has

not been reported in previous literature as to my knowledge.

1.4.3 Optimism, resources, and success

Optimists normally have more positive feelings and feel happier than pessimists in

various contexts. Due to the better coping strategies and better psychological

adjustments optimists have, and the resulting better health than pessimists, it is

plausible to infer that being optimistic can transform short-term optimistic tendencies

to a long-term approach of persistent goal pursuit and active coping strategies, which

endows optimists with more advantageous socio-economic resources and superior

opportunities for being successful than pessimists. Gould, Dieffenbach, and Moffett

(2002) interviewed 10 Olympic champions about their psychological characteristics.

They found that extraordinary athletic performance was characterized by higher than

average level of dispositional optimism and hope. Here next I will review some



related empirical studies in the literature illustrating the positive relationship between

optimism, resources, and success.

Previous studies demonstrate that students with higher levels of optimism

deal more easily with their first-year transition both socially and academically than

students scored lower in optimism. For example, a group of freshmen took a battery

of measures (including dispositional optimism, self-esteem, coping, depression,

perceived stress, and perceived social support) both at the beginning and at the end of

the starting semester (n = 89). Students with higher levels of dispositional optimism

experienced fewer increases in stress and depression, and greater increases in access

to social networks than pessimistic students over the first semester of college

(Brissette et al., 2002). Similarly, in a much larger sample of college freshmen (n =

2,189), L. S. Nes, Evans, and Segerstrom (2009) also found that optimistic students

had better psychological adjustment and motivation than pessimists in the period of

college transition. Students with a higher level of dispositional optimism were more

likely to return to school for the second year, with increased motivation and

decreased distress.

Similar results were found in studies involving attributional style in academic

backgrounds. For example, Peterson and Barrett (1987) reported that first-year

students with a positive explanatory style were more likely to have specific academic

goals and to utilize academic advising systems more efficiently, resulting in higher

grades on exams than students with a negative explanatory style. Benefits of an

optimistic explanatory style was expanded to athletic performance as well (Gordon,

2008).

Other studies have shown that optimists may also have better job

performance and higher income than pessimists. For instance, in a study conducted

within groups of insurance agents, Seligman and Schulman (1986) found that people

with a more positive explanatory style were more likely to keep their jobs after the

internship period, and tended to get a higher level of assessment on job performance.

Suzanne C. Segerstrom (2007) investigated the association between dispositional

optimism and several social resources in a group of law students. The 10-year



follow-up study found that students with higher levels of optimism before starting

school predicted higher income 10 years later. Further, both self-serving attributional

bias and dispositional optimism were found to be positively correlated with self-

confidence and forecast of future performance in a study with a group of MBA

students (Libby & Rennekamp, 2012).

One reason for the positive link between optimism and job performance may

be due to higher levels of career planning in optimists. Creed et al. (2002) found that

dispositional optimism was positively correlated with career exploration and career

planning. People who scored highly on the LOT-R produced more career-related

goals, and expressed more confidence about their career planning.

In addition, the benefits of being optimistic on social domain may also partly

account for optimists’ success in academic and career performance. MacLeod and

Conway (2005) reported that people with more positive expectations for the future

tended to have broader social networks. The longitudinal study described earlier also

demonstrated that increases in optimism were linked to developing larger social

networks across a 10-year period, indicating that optimism and social networks may

reinforce each other (Suzanne C. Segerstrom, 2007). Basically, optimists are

assumed to hold a better management in social relationships than pessimists (for a

review, see Carver & Scheier, 2014).

1.4.4 Optimism interventions included in positive psychology interventions

Psychologists and therapists have traditionally equated mental health with the

absence of mental illness. When a patient improved, he or she was taken to be

psychologically well. This view was fundamentally changed when positive

psychology was merged into mental health research and practice (Seligman, Steen,

Park, & Peterson, 2005). Previous research has shown that well-being can be

promoted by engaging in diverse positive activities, such as savouring (Bryant &

Veroff, 2007), practicing forgiveness (Reed & Enright, 2006), using signature

strengths (Linley, Nielsen, Wood, Gillett, & Biswas-Diener, 2010), and expressing

optimism and gratitude (Lyubomirsky, Dickerhoof, Boehm, & Sheldon, 2011). These



activities, so-called positive psychology interventions (PPI), aim to boost positive

emotions, thoughts, and behaviours, and other desirable consequences. Empirical

studies have indicated that these positive activities are effective for promoting well-

being and decreasing negative symptoms (for a review, see Sin, Della Porta, &

Lyubomirsky, 2011).

Even before the promotion of positive psychology by the American

Psychology Association, many different kinds of positive intervention methods had

been developed, though it is true that this trend has been greatly enhanced since

positive psychology has emerged. With the development of positive psychology

interventions, more and more controlled PPI designs began to explore their clinical

practice on people with mental illness, especially depression. A growing number of

positive psychology interventions have been tested on people with depressive

symptoms and those clinically diagnosed with depressive disorders.

The efficacy of specific positive perspectives has been proved in promoting

well-being and decreasing depressive disorders. A meta-analysis of 51 positive

psychology interventions (including optimism interventions) revealed that this form

of treatment is effective for improving well-being (r=0.29) and ameliorating

depressive symptoms (r=0.31). Findings suggested that clinicians should be

encouraged to incorporate positive psychology techniques into their clinical work,

particularly for treating depression. Also, delivering positive psychology

interventions as individual and group therapy and for relatively longer periods of

time is strongly suggested (Sin & Lyubomirsky, 2009).

Another review paper on positive psychology intervention research proposed

neural models for how such treatment might relieve depression, based on theory and

outcomes of research in social psychology, affective neuroscience and

psychopharmacology (Layous, Chancellor, Lyubomirsky, Wang, & Doraiswamy,

2011). For clinical depression treatments, some pioneering positive psychology

interventions, which consist of multiple positive-psychology based exercises, have

also been developed. For example, Seligman and colleagues (2006) proposed

positive psychotherapy (PPT) based on his new conceptualization of happiness and

previous positive psychology interventions in clinical practice.



Seligman and colleagues (2006) carried out a pilot study for testing the

efficacy of individual PPT. Thirty-two participants diagnosed with MDD (scores

more than 50 on the ASRS) were assigned to three groups: individual PPT group,

treatment as usual group (TAU), and TAU and antidepressant medications group

(TAUMED). For this study, PPT, which consisted of 14 sessions (including

optimism and hope interventions) during a period of 12 weeks, was administered to

address both positive and negative aspects of the clients. It showed that clients in the

PPT group reported greater well-being, more improvement in depressive symptoms,

and higher rates of remission, compared with clients in the other two groups. By

identifying and using the client’s character strengths, PPT established a balance

between promoting positive emotions and reducing negative depression. It was a

remarkable benefit for the clients to be taught positive social techniques, which

greatly promoted their consciousness of being kind, having gratitude and savouring

life.

Optimism interventions have been integrated with other positive activities in

most of previous practices, and have been tested singularly as well in some multi-

intervention studies. Research shows that optimism interventions are effective in

enhancing well-being and deducing negative emotions (Austenfeld, Paolo, & Stanton,

2006; Burton & King, 2004; Fosnaugh, Geers, & Wellman, 2009; Littman-Ovadia &

Nir, 2014; Meevissen, Peters, & Alberts, 2011).

1.4.5 Underlying mechanism: optimism and coping

It has long been believed that optimism may confer positive effects on psychological

and physical well-being (Carver et al., 2010; Nes & Segerstrom, 2006). The potential

mechanism underlying these benefits has been explored in numerous studies, the

majority of which proposed the importance of coping strategies. Coping is regarded

as a straightforward influence of optimism and pessimism regarding how people feel

when they encounter problems (Carver et al., 2010).

Theoretically, coping has been defined as “the cognitive and behavioural

efforts made to master, tolerate, or reduce external and internal demands and

conflicts among them” (Folkman & Lazarus, 1980). By this definition and the



differential nature of people, it is plausible to expect people are different in coping

with problems or stressful situations within their own environments. Additionally,

the widely accepted distinction in conceptualizing coping is between problem-

focused coping and emotion-focused coping (Folkman & Lazarus, 1980; Folkman &

Moskowitz, 2004), in which the former addresses external demands of stressors and

the latter addresses internal demands of problems. Another distinction conceptualizes

coping as approach coping (dealing with the demands of the stressors) and avoidance

coping (escaping from the demands of the stressors or emotions related to the

stressors) (Suls & Fletcher, 1985).

Theoretically, the construct of dispositional optimism stemmed from an

expectancy-value model in which behaviour embodies the pursuit of desired goals,

and a general self-regulatory model in which positive expectations arouse increased

effort to achieve desired goals (Carver & Scheier, 2001; Scheier & Carver, 1985).

This assumption is supported in empirical studies and shows that positive

expectancies lead to involvement and continued effort to attain desired goals,

whereas pessimistic expectancies lead to disengagement and reduced effort from

goal pursuit (L. S. Nes, Segerstrom, & Sephton, 2005). As a personality trait,

optimism could affect particular ways of thinking and behaving. It is reasonable to

expect that there is a potential mediating role of coping between optimism and

adjustment to specific situations.

Scheier and Carver (1985) reported their findings about the beneficial effects

of dispositional optimism on physical well-being, and proposed that these benefits

could be attributed to the increased likelihood of successful coping held by optimists

who normally take actions early when being confronted with problems. This claim

was supported by a study within a group of college students (Scheier, Weintraub, &

Carver, 1986). The authors found that optimists and pessimists differ in the strategies

they use to cope with stressful episodes. Compared with pessimistic participants,

optimistic subjects prefer problem-focused coping when they confront stressful

situations. The optimists seek social support and focus on positive aspects of the

stressful episodes. Comparatively, pessimists tend to use emotional-focused coping

and emphasize stressful feelings.



A number of studies further support the potential role of coping strategy for

mediating optimism and stress. In one study, undergraduates were asked to recall the

most stressful event they had experienced in the last month and complete a survey of

coping strategies relating to that event (Carver, Scheier, & Weintraub, 1989). The

authors found that dispositional optimism was positively correlated with active

problem-focused coping (r = .32). Billingsley et al. (1993) reported similar results in

their study, which was conducted with 82 college students over a period of four

weeks (r = .38 for Time 1 and r = .29 for Time 2). In another study with a larger

sample (420 undergraduates), dispositional optimism was also found to be positively

correlated with active coping strategy (r = .23) (Fontaine, Manstead, & Wagner,

1993). A meta-analysis of 56 studies revealed that the average weighted correlation

between dispositional optimism and coping strategies was .20 (Andersson, 1996).

Differences in coping strategies between optimists and pessimists have been

investigated in some studies with specific contexts. For instance, in one pioneering

study conducted within a group of cancer patients (Carver et al., 1993), 59 patients

who were diagnosed with early-stage breast cancer were interviewed to assess their

levels of optimism and coping strategies before and after their surgeries. It revealed

that optimistic patients initiated coping efforts before surgery and used different

coping strategies to deal with the crisis. Another study with a larger sample of 165

breast cancer patients reported similar results (Schou, Ekeberg, & Ruland, 2005).

High levels of dispositional optimism have been linked to positive coping styles in

some specific groups, such as women executives (Fry, 1995), cancer patients

(Horney et al., 2011; Llewellyn et al., 2013), HIV-infected patients (Rogers, Hansen,

Levy, Tate, & Sikkema, 2005), postnatal women (Rauch, Defever, Oetting, Graham-

Bermann, & Seng, 2013), and athletes (Chirivella, Checa, & Budzynska, 2013;

Nicholls, Polman, Levy, & Backhouse, 2008; Thompson, Gaudreau, Hoar, Hadd, &

Lelievre, 2008), and with particular backgrounds, including the work environment

(Strutton & Lumpkin, 1992) and posttraumatic situations (Prati & Pietrantoni, 2009).

Nes and Segerstrom (2006) investigated the relationship between

dispositional optimism and coping strategy in one meta-analysis (K = 50, N =

11.629). Both categories of coping distinctions (problem-focus versus emotion-focus,



and approach versus avoidance) were included. Analysis results showed that

dispositional optimism correlated positively with problem-focused coping (r = .13)

and approach coping (r = .17), and correlated negatively with emotion-focused

coping (r = -.08) and avoidance coping (r = -.21). It revealed that optimists are

inclined to eliminate, reduce, or handle stressors or related emotions when

confronting stressful situations, while pessimists seek to ignore, avoid, or escape

from stressors or emotions emerged. This stable coping tendency is especially

apparent for the distinction between approach and avoidance coping strategies.

The potential mediating role of coping between optimism and beneficial

results has been mainly restricted to dispositional optimism in previous literature to

my knowledge. There are few studies examining the potential mechanism

underpinning the benefits of explanatory style in the literature so far. Some

researchers, however, began to address this issue recently. For example, in a study

conducted with 205 adults, Sanjuan and Magallares (2014) found that attributional

style was positively correlated with active coping (r = .35) and negatively correlated

with avoidant coping (r = -.35). Structure model analysis indicated that coping

strategies mediated the relationship between attributional style and subjective well-

being.

1.5 Outline of the current research

1.5.1 Optimism in positive psychology

Though optimism has long been a focus of interest in the field of psychology, it has

been expanded exponentially since the initiating and rising of positive psychology.

The underlying assumption of positive psychology is that positive states or

traits are not necessarily the obverse of negative experiences and traits; and positive

emotions and behaviours are described by a completely separate psychological

process that functions via an isolated neural mechanism (Duckworth, Steen, &

Seligman, 2005). Positive psychology was proposed as ‘the scientific study of

positive experience and positive individual traits, ..., a field concerned with well-



being and optimal functioning…’ (Duckworth et al., 2005). On the basic level,

positive subjective experience in the past (e.g. life satisfaction), the present (e.g.

sensual pleasure), and in the future (e.g. optimism) are taken as important individual

levels in positive psychology (Seligman, 2002).

One reason I have focused on optimism emerges from the basic findings of

this trait in positive psychology. Positive psychology often focuses on well-being as

an outcome (Duckworth et al., 2005). It also focuses on resources for resilience, or

character strengths (Park, Peterson, & Seligman, 2004). Park et al. (2004) reported

that of 24 character strengths that he identified one, optimism, had the strongest link

to life satisfaction – one of three significant marks of well-being. Over the last 35

years, hundreds of cross-sectional and longitudinal studies have revealed that

optimism is positively associated with a host of psychological variables, such as self-

esteem, academic achievement, coping strategy, positive emotions, and perhaps most

importantly, predicts psychological and physical well-being both in the presence and

absence of stressors (Carver & Scheier, 2014; Carver et al., 2010; Forgeard &

Seligman, 2012; Scheier & Carver, 1992). Optimism seems to be a desirable

personality trait and individual variable, attracting more and more attention in the

field of positive psychology.

Another reason to focus on optimism came from some promising findings for

optimism interventions. Based on the widely-accepted correlations between

optimism and many other positive outcomes across individuals and contexts, positive

interventions in optimism have been designed to improve psychological well-being

by enhancing an individual’s optimistic expectations. Some optimism interventions

have been practiced in longitudinal experimental studies (Duckworth et al., 2005;

Seligman et al., 2006). In some of these studies, optimism interventions were

combined into the whole framework of positive psychotherapy (e.g. Seligman et al.,

2006; Seligman et al., 2005). In some other studies, optimism interventions were

taken as main therapy methods (e.g. Johnstone, Rooney, Hassan, & Kane, 2014;

Littman-Ovadia & Nir, 2014; Meevissen et al., 2011). Results of these studies

supported that optimism interventions were effective in increasing psychological

well-being and reducing negative emotions.



As described earlier, optimism has been conceptualized and measured in

different ways, among which dispositional optimism and optimistic explanatory

style are regarded as two main contrasting approaches (Carver et al., 2010; Forgeard

& Seligman, 2012). Though there are other psychological constructs proposed as

explanations for the optimistic thinking process, such as the cognitive model of hope

(Snyder et al., 1991), here in my research, optimism, if not specified, refers to the

two main approaches, dispositional optimism and explanatory style.

There are many promising aspects of optimism to be investigated and

explored. Over the last three and half years, my work focused mainly on two themes,

of which the first is to understand what optimism is and how we measure it, and the

second is to explore the possibility of optimism interventions on depressive

symptoms. The research described in the thesis consists of two main parts. Part I

incorporates measurement issues and conceptual ideas of optimism (from Chapter 2

to Chapter 6). Part II involves optimism interventions on depressive symptoms

(Chapter 7 and Chapter 8).

1.5.2 Part I measurement and concepts of optimism

In the first part of my study, I focus on some basic and important aspects of optimism,

including five points that concern measurement and concepts of explanatory style

and dispositional optimism.

First, I investigated the potential psychometric structure of the ASQ and the

LOT-R. As the most widely-accepted measure for explanatory style and for

dispositional optimism respectively, the ASQ and the LOT-R have been applied in

numerous studies. As mentioned earlier, the ASQ assigns subjects an optimistic or a

pessimistic explanatory style. An optimistic explanatory style consists of explaining

positive events as enduring, global and internally generated, while also explaining

negative events as unstable, specific, and externally caused (Forgeard & Seligman,

2012). If we are to understand the mechanism by which clinical and life outcomes

are influenced by explanatory style, it is important that we understand the structure

of the ASQ, decomposing the complex admixture of attributions, valences and events.



These components may have effects that are not apparent in a simple summing up of

positive and negative scores.

Similarly, though the LOT-R was originally supposed to measure a one-

dimensional bipolar construct of dispositional optimism (Scheier et al., 1994),

evidence from some studies indicates that the positively and negatively phrased items

in the measure split into two factors – dispositional optimism and dispositional

pessimism (e.g. Chang et al., 1997). It is important to address this issue before we

apply the LOT-R in our other studies.

Second, both explanatory style and dispositional optimism have been

assessed in their linkage to traditional personality traits, and most studies found that

optimism was positively correlated with Extraversion, and negatively correlated with

Neuroticism (e.g. Boland & Cappeliez, 1997). However, inconsistent results were

found in other studies. For example, Musgrave-Marquart, Bromley, and Dalley (1997)

reported that optimistic explanatory style was modestly correlated with

Conscientiousness but none of the other dimensions of the personality scale. Since

optimism is taken as relatively stable individual personality trait, it is important to

use traditional and well-established personality constructs as external criteria,

investigating the relationship between optimism and personality traits. So, I

examined correlations between two main approaches of optimism, explanatory style

and dispositional optimism, and the Five-Factor Model of personality (FFM; McCrae

& Costa, 1987).

Third, though optimism has been linked to well-being in previous studies and

both optimistic explanatory style and dispositional optimism have been identified as

positive factors in promoting well-being, few investigations have tested both

dispositional optimism and explanatory style in the research of psychological well-

being. Additionally, studies in which both explanatory style and dispositional

optimism are measured in the same sample have yielded inconsistent results on the

relationship between these two constructs. My study aimed to test a mediating model

in which dispositional optimism mediates the link between explanatory style and

psychological well-being.



Next, optimism-related research in recent years has been mainly conducted in

Westerners or English-speaking countries particularly, and it therefore may be less

valid for understanding the behaviours in population of other cultures. Since

examination of optimism across different cultural and ethnic groups is a crucial but

often neglected concern, the potential cultural differences between certain Easterners

(Mainland Chinese) and Westerners (White British) were investigated. I compared

the levels of optimism expression in these two ethnic groups, and explored cultural

indications of the results.

Finally, after examination of basic and fundamental issues in psychometric

structure and associations with personality and psychological well-being, I conducted

a pilot study on the basis of core concepts and measurement of attributional style.

Previous research has confirmed that people often give optimistically biased

attributions regarding themselves. However, it remains unclear what individuals

would do when they are explaining the same events for other people. I examined

attributional biases using new measures that are adapted from the standard ASQ.

1.5.3 Part II optimism interventions

Because of all the direct or indirect associations between optimism and personal and

social benefits, it is not surprising that optimism is reported to be relevant to clinical

psychology. Though positive psychology interventions have been applied in some

pioneering studies, very little systematic work has been done to investigate potential

advantageous effects of optimism interventions on psychotherapy applications in

concrete settings. How to convert the benefits of optimism to systematic and

effective interventions assisting pessimists to cope more actively with adversities in

their lives is still underexplored.

Optimism interventions applied in previous studies consisted of different

manipulation techniques, in which the Best Possible Self (BPS; Lyubomirsky et al.,

2011), and the self-administered optimism training (SOT; Fresco, Moore, Walt, &

Craighead, 2009) have been developed on the theoretic basis of dispositional

optimism and explanatory style respectively. Applications of these two optimism

manipulations in empirical studies have yielded results confirming the positive



effects of optimism interventions on enhancing well-being. However, no research

including both these optimism interventions has been conducted so far to my

knowledge.

On the basis of previous findings that both these optimism techniques are

effective in promoting psychological well-being and reducing depressive symptoms,

in the second part of my research, I designed and conducted two studies to test the

advantages of optimism interventions in reducing dysphoria. Two different optimism

manipulations were adapted from the BPS and SOT respectively. These two

optimism intervention strategies were applied in two experiments in two

undergraduate samples, aiming to investigate the beneficial effect of optimism

interventions on depressive symptoms.

1.5.4 Measures

Eight measures in total were involved in my research.

1.5.4.1 The Attibutional Style Questionnaire (ASQ)

The original English version of the ASQ (Peterson et al., 1982) was used to measure

explanatory style of the British students. Attributional Style of Mainland Chinese

participants was measured using a Mainland Chinese version of the ASQ (Zhang,

2006). The original English version of the ASQ was obtained from Dr Seligman,

who granted permission to use this test for research purposes.

Just as the original English version of ASQ, the Chinese ASQ is composed of

12 different hypothetical situations, consisting of 6 positive events (e.g., “You do a

project that is highly appraised”) and 6 negative events (e.g., “You have been

looking for a job unsuccessfully for some time”). Each of these 12 different

hypothetical situations is followed by a series of 4 questions which are arranged in

the same order. The first question following each situation asks for the one major

cause of the situation. This question is not used in scoring and simply serves as an

aid to better answer the remaining questions. The remaining three questions are

arranged in the same order for each situation and measure three different dimensions.

The second question following each situation measures whether the subject’s



response is internal or external (e.g. “is the cause of your unsuccessful job search due

to something about you or to something about other people or circumstances”). The

third question following each situation measures whether the subject’s response is

stable or unstable (e.g. “in the future when looking for a job, will this cause again be

present”). The fourth question following each situation measures whether the

subject’s response is global or specific (e.g. “is the cause something that just

influences looking for a job or does it also influences other areas of your life”).

For each response, subjects marked an answer in the range of 1 to 7. (for

internality vs. externality dimension, from ‘Totally due to other people or

circumstance’ to ‘Totally due to me’; for stability vs. instability dimension, from

‘Will never again be present’ to ‘Will always be present’; for globality vs. specificity

dimension, from ‘Influence just this particular situation’ to ‘Influence all situations

in my life’). For positive events, a score of 1 is the lowest or worst possible score,

whereas a score of 7 is the highest or best possible score. Conversely, for negative

events, a score of 1 is the highest or best possible score, and a score of 7 is the lowest,

or worst possible score. Reliabilities for the original English version of the ASQ

were reported as = .50 for Internal Positive, = .58 for Stable Positive, = .44 for

Global Positive, = .46 for Internal Negative, = .59 for Stable Negative, and

= .69 for Global Negative (Peterson et al., 1982). Reliabilities for the original

Mainland Chinese version of the ASQ were reported as > .77 (apart from

internality, where = .49) (Zhang, 2006).

Traditionally, the scale produces scores for the explanation along the theme

of positive and negative events (Peterson et al., 1982). As a result, composite

attributional styles were calculated separately for positive and negative events.

Higher scores for positive events and lower scores for negative events on any area

demonstrate a more “optimistic” attributional style for that domain, i.e., more

external, temporary and specific for bad events, and more internal, stable and global

for good events. Generally, the ASQ scoring produces three composite scores and six

scores of the individual dimension measures based on participants’ responses to the

scale items. The three composite scores are Composite Negative (CoNeg, CN, or



ASQ Negative), Composite Positive (CoPos, CP, or ASQ Positive), and Composite

Positive minus Composite Negative (CPCN or ASQ Total). Here the CPCN scoring

is theoretically based on the belief that an optimistic explanatory style is explicit

when people make attributions for both positive and negative events they encounter

in life.

In some cases, two other composite scores, Hopelessness and Hopefulness,

are also produced separately for negative and positive events, based on some

research results that the stability and globality factors correlated strongly, with

internality-externality being more independent (Higgins et al., 1999). The six

individual dimension scores are Internal Negative, Stable Negative, Global Negative,

Internal Positive, Stable Positive, and Global Positive. This scoring method was

applied in almost all previous studies dealing with explanatory style in the literature.

1.5.4.2 The Attributional Style Questionnaire – Other (ASQ – Other)

Attributional Style for others was measured using an adapted Chinese version of the

ASQ, differing in that subjects are asked to imagine the event occurring to a fictional

character “Wang Chen”, described as being a healthy undergraduate of normal

intelligence. The same events, instructions to generate causes, and ratings scales

were used as in the ASQ.

As in the standard ASQ, 12 events, 6 positive and 6 negative, were divided

across the domains of achievement and affiliation in the ASQ-Other. Respondents

were asked to generate a likely cause for such an event, and, subsequently, to rate

these causes on the following three characteristics: Internal versus external causation

(e.g. “is the cause of Wang’s unsuccessful job search due to something about Wang

Chen or to something about other people or circumstance”), stability versus

instability (e.g. “in the future when looking for a job, will this cause again be present

for Wang Chen”), and specificity versus global applicability (e.g. “is the cause

something that just influences looking for a job or does it also influence other areas

of Wang Chen’s life”).

All responses are on the same 7-point scale (for internality vs. externality

dimension, from ‘Totally due to other people or circumstance’ to ‘Totally due to



Wang Chen’; for stability vs. instability dimension, from ‘Will never again be

present for Wang Chen’ to ‘Will always be present for Wang Chen’; for globality vs.

specificity dimension, from ‘Influence just this particular situation in Wang Chen’s

life’ to ‘Influence all situations in Wang Chen’s life’).

1.5.4.3 The Attributional Style Questionnaire – General (ASQ – General)

Attributional Style for general situations was measured using an adapted Chinese

version of the ASQ, differing in that subjects are asked to imagine the event

occurring for all people on average, not just the participants themselves. The same

events, instructions to generate causes, and ratings scales were used as in the ASQ.

As in the original, 12 events, 6 positive and 6 negative were included.

Respondents were asked to generate a likely cause for such an event, and,

subsequently, to rate these causes on the same three characteristics as above: Internal

versus external causation, stability versus instability, and specificity versus global

applicability. All responses were rated on the same 7-point scale.

1.5.4.4 The Life Orientation Test-Revised (LOT-R)

The original English version of the Life Orientation Test-Revised (LOT - R; Scheier

et al., 1994) was used to measure dispositional optimism in the British sample. A

Mainland Chinese version of Life Orientation Test-Revised (CLOT-R; Lai et al.,

1998) was used to measure dispositional optimism of the Mainland Chinese students.

The LOT-R is a brief modified version of the original Life Orientation Test

(LOT; Scheier & Carver, 1985) and has been found to correlate 0.95 with the LOT

(see Scheier et al., 1994). Support for the construct validity of the LOT-R has been

reported in Scheier et al. (1994). Just as in the original English version of LOT-R,

the CLOT-R comprises three positively phrased items (e.g. “In uncertain times, I

usually expect the best”), three negatively phrased items (e.g. “I hardly ever expect

things to go my way”), and four filler items. The psychometric properties of the

Mainland Chinese LOT-R were reported by Lai et al. (1998) as = .70. For all items

of both the English and Chinese versions, see the Appendix. Respondents are

directed to assess the extent to which they agree with each of the 10 items on a 5-



point scale (4 = strongly agree, 3 = agree, 2 - neutral, 1 = disagree, and 0 = strongly

disagree).

1.5.4.5 The Ryff Scales of Psychological Well-being (RSPW)

Psychological well-being was measured with a Chinese version of the Ryff Scales of

Psychological Well-being (Chen, 2010). The original English version of the RSPW

and this Chinese version of the RSPW were obtained from Dr Ryff, who granted

permission to use these tests for research purposes.

The Chinese version of the RSPW consisted of nine items for each of the six

dimensions: Self-Acceptance (e.g. “I made some mistakes in the past, but I feel that

all in all everything has worked out for the best”), Positive Relationships With

Others (e.g. “Maintaining close relationships has been difficult and frustrating for

me”), Personal Growth (e.g. “When I think about it, I haven't really improved much

as a person over the years”), Environmental Mastery (e.g. “The demands of everyday

life often get me down”), Autonomy (e.g. “I am not afraid to voice my opinions, even

when they are in opposition to the opinions of most people”), and Purpose in Life

(e.g. “I enjoy making plans for the future and working to make them a reality”).

Items were rated on a 6-point Likert scale (1 = strongly disagree; 6 = strongly agree).

1.5.4.6 The Revised NEO Personality Inventory (NEO-PI-R)

Personality was measured with a Chinese version of the Revised NEO Personality

Inventory (Yang et al., 1999).

Just as in the original English version of the NEO-PI-R (Costa & McCrae,

1992), the Chinese version contains the same 240 items with five domain scales

assessing the five broad personality traits of the Five-Factor Model of personality

(FFM; McCrae & Costa, 1987): Neuroticism (e.g. “I often get angry at the way

people treat me”), Extraversion (e.g. “I don’t get much pleasure form chatting with

people”), Openness to Experience (e.g. “I don't like to waste my time daydreaming”),

Agreeableness (e.g. “I believe that most people will take advantage of you if you let

them”), and Conscientiousness (e.g. “Over the years I’ve done some pretty stupid

things”). Items were rated on a 5-point Likert scale (1 = strongly disagree; 5 =



strongly agree). Reliabilities for the Chinese version of NEO-PI-R scale were

reported ranging from .77 to .91 (Yang et al., 1999).

1.5.4.7 The Beck Depression Inventory (BDI)

A Chinese version of the Beck Depression Inventory (BDI; Chan & Tsoi, 1984),

which was translated from the original version of the BDI (Shek, 1990) was used to

measure depression. Chan and Tsoi (1984) reported the split-half reliability

coefficient between odd and even items was .62 (p < .05), and test-retest reliability

was .72 (p < .05).

The BDI is a 21-item, self-report measure that broadly assesse the symptoms

of depression including affective (e.g. “I feel quite guilty most of the time”),

cognitive (e.g. “I feel my future is hopeless and will only get worse”), somatic (e.g. “I

have lost more than ten pounds”), and motivational components (e.g. “I blame myself

for everything bad that happens”), as well as suicidal wishes (e.g. “I would like to

kill myself”). Each item in the BDI describes a specific behavioural manifestation of

depression (such as loss of appetite or somatic problem), and each symptom item

consists of several statements that range from neutral to severe forms of symptoms.

Assignment of a consistent weighted score of 0, 1, 2, and 3 was used for each item.

Admittedly, it seems that there are some overlap between BDI and LOT-R,

since they both ask about bad expectations about the future. However, as the

perspective of “dispositional optimism” originated theoretically from the expectancy-

value model and put much emphasis on confidence or doubt pertaining to

life, optimism and pessimism are broad, generalized versions of expectations to

future life, rather than to just a specific narrow context.

1.5.4.8 The Satisfaction with Life Scale (SWLS)

Subjective well-being was measured with an on-line based Chinese version of the

Satisfaction with Life Scale (SWLS; Chen & Zhang, 2004), which was translated

from the original English version (Diener et al., 1985).



The SWLS is a five-item scale that measures general life satisfaction. It

includes items such as ‘If I could live my life over, I would change almost nothing.’

Responses are on a 7-point scale (1: strongly disagree to 7: strongly agree).

1.5.5 Participants

Sample 1

A total of 452 participants were included in sample 1, of which 267 undergraduates

were recruited from Jinan University (JU), and 185 undergraduates were recruited

from China Youth University for Political Science (CYUPS). Both these universities

are located in Mainland China. All participants were native Chinese speakers.

Participants in sample 1 completed the ASQ, the ASQ-Other, the LOT-R, the RPWB,

and the NEO-PI-R.

In sample 1, there were 133 males (mean age = 20.70, SD = 1.30) and 319

females (mean age = 20.46, SD = 1.24). All participants took part in the present

study on a voluntary and anonymous basis.

Sample 2

A total sample of 232 participants was recruited from the CYUPS (different subjects

from sample 1). The participants were aged 17-21 years (mean age=18.76 years,

SD=0.89); 97 males, 135 females. All participants in sample 2 took part in the

present study on a voluntary and anonymous basis. All participants were native

Chinese speakers. Participants in sample 2 completed two measures, the ASQ and

the LOT-R. All participants took part in the present study on a voluntary and

anonymous basis.

Sample 3

A total sample of 205 White British participants were recruited among students

enrolled in a social science course in Edinburgh Napier University; 46 males and 159

females (mean age=20.10 years, SD=0.87). All participants were native English

speakers. All participants in sample 3 took part in the present study on a voluntary

and anonymous basis. Participants in sample 3 completed two measures, the ASQ



and the LOT-R. All participants took part in the present study on a voluntary and

anonymous basis.

Sample 4

A total sample of 117 participants was recruited from Jinan University (different

subjects from sample 1). The participants were aged 18-23 years (mean age=19.79

years, SD=1.11); 25 males, 92 females. All participants in sample 4 took part in the

present study on a voluntary and anonymous basis. All participants were native

Chinese speakers. Participants in sample 4 completed the ASQ-General. All

participants took part in the present study on a voluntary and anonymous basis.

Sample 5

Fifty-two freshmen (22 males and 30 females) with depressive symptoms were

recruited from the CYUPS. All participants were native Chinese speakers with ages

ranging from 17 to 21. All participants in sample 5 took part in the present study on a

voluntary basis. The 52 participants were randomly divided into one of the two

conditions: an experimental group (n = 26) and a no-treatment control group (n = 26).

Not all participants completed the whole procedure. Three participants dropped out

of the intervention group and two dropped out of the control group. As a result, there

were 23 participants in intervention group and 24 participants in the control group

available for the final data analysis (Mage = 18.83, SD = 0.84) (19 males and 28

females).

Participants in sample 5 completed four measures, the BDI, the ASQ, the

LOT-R, and the SWLS. All participants took part in the present study on a voluntary

and anonymous basis.

Sample 6

Sixty-eight first-year university students (30 males and 38 females) were recruited

from the CYUPS (different subjects from sample 5). All participants were native

Chinese speakers with ages ranging from 17 to 21. All participants in sample 6 took

part in the present study on a voluntary basis.



The 68 participants were randomly divided into one of the two conditions: an

experimental group (n = 34) and a ‘placebo’-treatment control group (n = 34). Not all

participants completed the whole procedure. Four participants dropped out of the

intervention group and five dropped out of the control group. As a result, there were

30 participants in intervention group and 29 participants in the control group

available for the final data analysis (Mage = 19.03, SD = 0.74) (27 males and 32

females).

Participants in sample 6 completed four measures, the BDI, the ASQ, the

LOT-R, and the SWLS.

Data collected from sample 1 and sample 2 was used in the study

investigating psychometric constructs of the ASQ and the LOT-R, relationship

between optimism and psychological well-being, relationship between optimism and

personality, and exploration of attributional bias. Data collected from participants in

sample 2 and sample 3 were applied in the cross-cultural study of attributional style

and dispositional optimism. The study of attributional bias involved part of data

collected from sample 1 and data collected from sample 4. Participants in sample 5

and sample 6 were recruited to examine intervention effects of optimism

manipulations.


Chapter 2: The psychometric construct of optimism 50

Chapter 2: The psychometric construct of optimism

As the most widely-applied measures for explanatory style and dispositional

optimism respectively, the ASQ and the LOT-R have been psychometrically

analysed in a number of studies since they were originally developed. In addition to

controversial results and conclusions, previous studies in examining the

psychometric constructs of these two measures have mostly been conducted in

industrial countries. Investigating basic constructs of the ASQ and the LOT-R in

Eastern cultural backgrounds has theoretical and empirical importance.

2.1 The psychometric construct of the ASQ

Regarding the psychometric construct of the ASQ, I set out to accomplish two main

goals. First I wished to examine the structure of the ASQ using structural equation

modelling of attributions for positive and negative events simultaneously. Second, I

wished to test the role of cognitive style (such as global versus local explanations)

that might play a role over and above explanatory bias. For instance, attributions of

instability may apply to both positive and negative events. The literature motivating

these aims is reviewed below.

2.1.1 Myths about attributional style

Attributional style has been developed from the original two-factor structure to the

current widely accepted construct of three dimensions. Originally, two basic factors

of casual explanations for actions – internality, a factor “with the person”, which

occurs when an individual blames him or herself for a problem, and externality, a

factor “within the environment”, when one blames something outside of oneself,

were differentiated by Heider (1958). This notion of internality and externality was

supported by Weiner (1974), who developed stability – the consistency of the cause

– as another attributional component. Differentiation between stability and

instability depends on whether the cause is taken as everlasting or as fleeting.

Later on, globality, which is linked to the prediction of recurrence of the same

cause in other situations, was developed as a newly-applied notion of attributional



factor (Abramson et al., 1978). As a result, a three-dimensional model, which

incorporated dimensions of internality, stability, and globality, was put forward by

Abramson et al. (1978). Here in their opinion, internality and stability have basically

the same meaning as the two components identified above by Heider (1958) and

Weiner (1974). Thus far, these three dimensions, internal versus external, stable

versus unstable, and global versus specific, have been combined to form the three-

dimensional model of explanatory style. And the Attributional Style Questionnaire

(ASQ; Peterson et al., 1982) was developed on the basis of these three-dimensional

model of casual explanations.

As mentioned earlier, the ASQ assigns subjects an optimistic or a pessimistic

explanatory style. An optimistic explanatory style consists of explaining positive

events as enduring, global and internally generated, while also explaining negative

events as unstable, specific, and externally caused (Forgeard & Seligman, 2012). If

we are to understand the mechanism by which clinical and life outcomes are

influenced by explanatory style, it is important that we understand the structure of

the ASQ, decomposing the complex admixture of attributions, valences and events.

These components may have effects that are not apparent in a simple summing up of

positive and negative scores.

Within attributional models of depression, the attributions are seen to cause

heavy distinct behavioural consequences. For instance, low self-esteem is agreed to

be linked with internal attributions regarding negative events, while chronic

depression is suggested to result from stable attributions for negative events (Haugen

& Lund, 1998; Peterson et al., 1982). In this learned helplessness model, depression

emerges as a consequence of experience with uncontrollable negative events

(Abramson et al., 1978). Concept of attributional style however also predicts that the

three types of explanation are correlated each other within at least within each

valence. This is shown in graphically in Figure 2.1.



Figure 2.1: Proposed Model of attributional style based on learned helplessness

theory of responses to experience of negative events.

Research based on this model has resolved in findings that are somewhat

counterintuitive. The earliest data on this question was collected by Peterson et al.

(1982). They reported that attributions for positive events and attributions for

negative events were essentially uncorrelated (r = .02). This lack of correlation

between explanatory styles for positive and negative events has been found in other

work. For instance, P.J. Corr and J.A. Gray (1996) investigated the factor structure of

the ASQ in two independent samples using Varimax rotated principal components

analysis. They found that positive and negative explanatory styles were independent.

In addition, whereas for negative events, internality ratings were largely independent

of stability and globality ratings, for positive events these three dimensions formed a

single factor, suggesting that explanations for positive and negative events might

have different structures. The study of Bunce and Peterson (1997) also revealed that

there is no correlation between explanations for positive and negative events. This

independence was reported for ASQ composite score and the internality dimension

as well.



Subsequent studies have used larger samples, and incorporated confirmatory

structural equation modelling (SEM), allowing a better understanding of the structure

of attributions by contrasting competing theoretical models. For instance, Higgins et

al. (1999) reported a confirmatory factor analysis of the ASQ identifying three-

correlated factors in over 1,000 subjects. This model fitted well (RMSEA = .02) for

negative event attributions and for positive events as well (RMSEA = .02).

Consistent with several other previous studies, the stability and globality factors

correlated strongly (r = .61 for negative events, r = .67 for positive events), with

internality-externality being more independent of the globality (r = .35 for negative

events, r = .28 for positive events). Though different patterns appeared for negative

and positive events regarding the correlation between internality and stability factors

(r = .20 and r = .55 respectively).

The next major advance in modelling attributional style was the realization that,

because subjects are generating multiple responses to each event, analyses must

incorporate multi-method analytic strategies. This is an important innovation, as

misleading results can arise in analyses of data generated from multiple correlated

responses based on each item (as is true in the ASQ where all three attributions are

samples for each event).

Using a multi-trait multi-method (MTMM) model, Hewitt et al. (2004) found

that the three-factor structure of attributional style still provided a good account of

responses to negative events in terms of correlated latent factors of internality-

externality, stability-instability, and globality-locality. Contrasting, however, with

previous studies, and reflecting the importance of correct modelling of the multiple

assessments of each event, this model indicated higher correlations between

internality and the other two factors (r = .52 for internality and stability and .45

between internality and globality). Here only negative event attributions were tested

in this study.

The possibility of modelling both positive and negative event attributions jointly

raises the possibility of addressing two questions. First, such data can establish

whether attributions regarding the causes of positive events and negative events are



negatively correlated i.e., do individuals giving optimistic explanations for positive

events tend to give optimistic explanations for negative events?

Secondly, a very different model of the ASQ and of attributions can be posed

and tested. Rather than clustering around event valences to create an attributional

style in which good and bad events are attributed to different types of causes, instead,

subjects may have cognitive styles which apply independent of event valence, and

these style factors may account for a preponderance of variance in the ASQ. This is

shown graphically in Figure 2.2.

Figure 2.2: Proposed Model of Attributions in terms of valence-independent

cognitive styles, rather than valenced biases.

As shown in Figure 2.2, a cognitive style model predicts that the tendency to

apply global-local, internal-external and stable-unstable explanations to events may

be independent of event valence: The same person who tends to ascribe, say, an

internal cause to negative events may apply a similar internal explanation to positive

events in their lives. It is, therefore, important to distinguish between cognitive style

models, which would apply to events independent of valence, versus affect-linked

attributional style models.



To summarize the findings to date, it is clear that adequate analyses of the

structure of the ASQ require use of structural equation modelling and, in particular,

of multi-trait multi-method modelling to account for the repeated entry of events into

explanations (Campbell & Fiske, 1959; Hewitt et al., 2004). For negative events,

research confirms a three-correlated factor structure. However no study in which

both positive and negative events examined jointly have been conducted within

models controlling for correlated event structure. This leaves the structure of the full

ASQ unclear. In addition, a majority of studies to date have been conducted in

Western samples, and it is not known whether the structure of explanatory style is

invariant across culture.

After having examined relevant findings in current literature, I next outline in

detail the two major research questions explored in the present study.

The first analyses sought to replicate the three correlated factor structure for

negative events reported by Hewitt et al. (2004) using the MTMM model and the

similar factor structures for positive events revealed by Higgins et al. (1999). These

analyses can confirm (or disconfirm) that correlated factors of globality, stability,

and internality emerge for both kinds of event. However, analyses of the different

event valences in separate models miss the opportunity to test competing models

incorporating attributions for the causes of both positive and negative events. Full

data from positive and negative sections of the ASQ also allow testing a second

important question; that of disentangling cognitive styles from optimistic and

pessimistic attributions. It is to resolve these two questions that we turn next.

Data on attributions about both positive and negative events offer the

opportunity to test whether the three attribution factors emerging for each event type

are the same across events: That is whether globality for positive events is identical

to the factor influencing globality ratings for negative events, and likewise for

locality and internality as shown in Figure 2.2. To the extent that cognitive styles

have important influences on responses, people’s explanations of events will reveal

coherent attributional styles for events independent of event-valence (Rotter, 1966),

rather than explanations driven by experience with valence-specific outcomes



(Peterson et al., 1982). Of course both cognitive styles and valence-specific

optimistic explanatory style factors may exist. This combined model is shown in

Figure 2.3.3.

Figure 2.3: Combined framework for testing contrasting models of attributional Style.

Note: Explanatory style Models predict strong effects of valenced explanatory styles

(negative event explanations & positive event explanations). By contrast, cognitive

style Models predict large influences of internal –external, global – local & stable –

unstable processing, biases independent of event valence.

Figure 2.3.3 lays out the full complexity of analytic outcomes tested here. As can

be seen, six types of item response emerge from the ASQ: three attributions for each



of two event valences. These six response types are potentially accounted for by

three cognitive styles (upper portion of Figure 2.3.3), and/or by two valence bias

factors (lower portion of Figure 2.3.3). In addition, the three cognitive style factors

may correlate or be independent of one another, likewise, negative event

explanations may be negatively correlated with positive event explanations, or be

uncorrelated.

Importantly, if explanatory biases for positive events and for negative events are

uncorrelated, then the description of individuals as having either an optimistic or

pessimistic explanatory style will be based on a composite of causes, and most

individuals will have mixed biases. Alongside this, most people, if the cognitive style

factors are influential on attributions, will tend to generate the same kinds of

explanation for both positive and for negative events. And, the personal cognitive

style which is predicted to be depressogenic (Abramson et al., 1978), will,

paradoxically, be associated with a self-enhancing explanatory style for positive

events. To this extent, a notion of positive or negative attributional style would not

be applicable to most individuals.

Analyses and analysis techniques

We first replicated the model for negative events, and then extend this work to model

positive events. Finally, in the second section of the analyses, we model both positive

events and negative events simultaneously, testing the attributional style model, in

which attributions regarding positive and negative events are clustered. We tested

also if these clusters are correlated or not. These are contrasted with models in which

attributions are driven instead by differences in cognitive style, independent of event

valence, i.e., a tendency to ascribe events to local or stable causes, independent of

whether they are positive or negative. Following this work, a second study is

reported, replicating the proposed and confirmed joint model from study one in an

independent sample.

All data were analysed at the item level. All variables were approximately

normal. Given the 1-7 response scale for each item, data were analysed as continuous



(Rhemtulla, Brosseau-Liard, & Savalei, 2012). Models using polychromic input

rendered highly similar solutions and fits. Correlations among item responses were

used to estimate parameters in a confirmatory factor analysis framework, comparing

proposed theoretical models, as described above. Final models were permitted to

include explicit exploratory modifications where necessary (all modifications are

noted explicitly). Modelling was undertaken using OpenMx (Boker et al., 2011;

Boker et al., 2013) under R (R Core Team, 2012). All analyses took advantage or

raw data supporting estimation of models using full information maximum likelihood

estimation.

The adequacy of model fit was assessed using the comparative fit index (CFI),

Tucker-Lewis index (TLI) and the Root Mean Square Error of Approximation

(RMSEA). For CFI and TLI, values > 0.95 were taken as indicating acceptable fit

(Hu & Bentler, 1999). For the RMSEA, values of < .05 indicated acceptable fit (C. Y.

Yu, 2002). Akaike Information Criterion (AIC) and Bayesian Information Criterion

(BIC) are reported to aid model comparison.

2.1.2 Samples and instruments

Samples

There are two independent samples were included in this study (for detail of the two

samples, see 1.5.4 of Chapter 1). Sample 1 was involved in constructing and testing

the proposed model. Sample 2 was used to replicate the model. No subjects from the

replication study participated in the initial modelling analysis.

Instruments

Attributional style was assessed using the Chinese ASQ (Zhang, 2006). Composite

attributional styles were calculated separately for positive and negative events

separately. Higher scores for positive events and a lower score for negative events on

any area demonstrates a more “optimistic” attributional style for that domain, i.e.,

more external, temporary and specific for bad events, and more internal, stable and

global for good events.



Reliabilities (Cronbach’s α) were acceptable 0.84 for the total and, for

positive events 0.84; for negative events .77; for internality, .65; for stability, .76;

and .80 for globality.

Procedure

Participants were tested in groups of 30 to 50 by their lecturer. Each lecturer was

trained on the administration of the task. After detailed instructions were provided,

participants completed the paper-and-pencil questionnaires.

2.1.3 Testing models of causal attributions for positive and negative events

A total of 452 participants in sample 1 were involved in this testing.

Table 2.1 shows the descriptive statistics. Reliabilities were acceptable. No

significant gender differences emerged and the data were pooled across sex in

subsequent analyses.

Measures Means SD Cronbach’s Alpha

Negative Events 12.9 1.78 0.84

Internal Negative 4.45 0.67 0.49

Stable Negative 4.33 0.85 0.73

Global Negative 4.12 0.9 0.73

Positive Events 15.28 1.91 0.77

Internal Positive 5.03 0.7 0.65

Stable Positive 5.36 0.78 0.75

Global Positive 4.9 0.85 0.71

ASQ Total 2.38 2.17 0.84

Table 2.1: Means, SDs and Cronbach’s Alpha for the ASQ scales.

Note: Means for ASQ dimensions are on a scale ranging from 1 to 7. (n = 452)



2.1.4 Structural equation modeling

I first tested the hypothesis that the structure of explanations for the causes of

negative events reflects three factors of internality, stability and globality which are

correlated. As in Hewitt et al. (2004), method (event) variance was accommodated

using an MTMM structure. Hewitt et al. (2004) fit correlated factor models. Here I fit

both this and the (statistically similar but theoretically distinct) higher-order model in

keeping with the modelling to be undertaken below. Fit for both types of model is

identical, and the correlated factor correlations are reported. This model is shown in

Figure 2.4. For clarity, this correlated method variance is not shown on the figure.

Figure 2.4: Well-fitting 3-factor model of attributional style for negative events.

The base model without modifications fitted reasonably well (χ² (96) = 212.32, p

< .001; CFI = 0.94; TLI = 0.92; RMSEA = 0.044). Three modifications improved fit

(χ² (3) = 47.1, p<.001) by all criteria (χ² (93) = 165.25, p <.001; CFI = 0.97; TLI =

0.95; RMSEA = 0.033). The new paths all had loadings of.27 or below suggesting



the deviation of reality from the theoretical model is minor (see Figure 2.4). In a

correlated factor model, stability and globality correlated .47, internality and

globality had an r of .39, and internality and stability factors correlated = .20.

Thus, as previously reported by Hewitt et al. (2004), a model of causal

attributions for negative events in terms of three correlated factors of globality,

stability, and internality adequately accounted for responses to these negative events

in the ASQ. We next turned to see if this model would fit well for positive events.

A model for positive events was constructed in the same fashion as the baseline

model for negative events (see Figure 2.5). Fit measures for this model indicated

excellent fit between model and data (χ2

(96) = 152.48, p < 0.001; CFI = 0.98, TLI =

0.98, RMSEA = .027). No modifications were needed from base model. In the

correlated factor model stability and globality correlated .57, internality and

globality .48 and internality and stability .62: Considerably higher than was the case

for negative events.

Figure 2.5: Well-fitting 3-factor model of attributional style for positive events.



As a result, as previously reported by Higgins et al. (1999), a model of causal

attributions for positive events in terms of three correlated factors of globality,

stability, and internality adequately accounted for responses to these positive events

in the ASQ.

Analyses of separate ASQ positive events and ASQ negative events, then,

indicated that these scales were well accounted for by three correlated factors of

internality, stability, and globality. As can be seen in Figures 2.4 and 2.5, correlations

between the three dimensions were high and significant, especially for negative

events, where globality effectively defined the common factor.

I next moved on to construct models of both positive and negative ASQ events,

jointly testing the competing models outlined in the introduction and shown in Figure

2.3.

Joint modelling of attributions of causality for positive and negative events

The sequence and fit statistics of all joint models tested are laid out in Table 2.2.

I first tested a model accounting for positive and negative event attribution in

terms of just two negatively correlated factors of negative and positive event

attributions (See Figure 2.1 and Figure 2.3 lower section Figure 2.3). This fitted

poorly (χ² (521) = 1972.74, p < .001; CFI = 0.68; TLI = 0.64; RMSEA = 0.075; AIC

= 2190.74; BIC = 2639.13) (see Table 2.2). I next modified this model setting the

latent factors for positive and negative event attributions to be uncorrelated. This

model fitted better than the first, but remained less than adequate (χ² (515) = 1058.89,

p < .001; CFI = 0.67; TLI = 0.63; RMSEA = 0.076; AIC = 2232.34; BIC = 2676.72)

(see Table 2.2).

I next tested a model accounting for the data in terms of three cognitive styles,

i.e., in terms of tendencies to attribute global or local or stable causes to events,

irrespective of their valence. This model was constructed by creating three

uncorrelated latent variables: An Internal Style factor, with loadings from internality

attributions for both positive and negative events, and similar Stability-Style and



Globality-Style factors, also loading from their respective attributes across the two

event valences. This model fitted poorly (χ² (522) = 1509.05, p < .001; CFI = 0.79;

TLI = 0.76; RMSEA = 0.061; AIC = 1725.05; BIC = 2169.33) (see Table 3.2). We

therefore moved to a correlated cognitive styles model. This improved fit but was

still not adequate (χ² (519) = 1375.91, p < .001; CFI = 0.82; TLI = 0.79; RMSEA =

0.057; AIC = 1597.91; BIC = 2054.53) (see Table 2.2). Next the preferred model

containing both cognitive and explanatory style factors was tested.

Joint Models χ² /df CFI TLI RMSEA AIC BIC

Model 1 - correlated negative and

positive event explanations 3.79 0.68 0.64 0.075 2190.74 2639.13

Model 2 - uncorrelated negative

and positive event explanations 3.86 0.67 0.63 0.076 2232.34 2676.72

Model 3 - uncorrelated cognitive

styles 2.89 0.79 0.76 0.061 1725.05 2169.33

Model 4 - correlated cognitive

styles 2.65 0.82 0.79 0.057 1597.91 2054.53

Model 5 - “3-cognitive styles + 2-

explanatory styles” in study 1 (see

Figure 3.6)

1.39 0.97 0.96 0.025 982.74 1690.29

Model 6 – Replication of Model 5

in independent data 1.38 0.97 0.96 0.025 984.61 1695.13

Model 7 – Replication in the

combined data set (see Figure 3.7) 1.52 0.97 0.96 0.024 1044.19 1822.99

Table 2.2: Fit statistics for Attributional Style.

Note: CFI = the comparative fit index; TLI = Tucker-Lewis index; RMSEA = Root

Mean Square Error of Approximation. AIC = Akaike Information Criterion. BIC =

Bayesian Information Criterion. Preferred model (Model 5) in Bold.

Results supported the predicted model (χ² (483) = 845.42, p <.001; CFI = 0.93;

TLI = 0.91; RMSEA = 0.037). Modifications were suggested yielding good model fit



(χ² (458) = 634.34, p <.001; CFI = 0.97; TLI = 0.96; RMSEA = 0.025; AIC = 982.74;

BIC = 1690.29) (see Table 2.2 and Figure 2.6). The new paths all had loadings of .23

or below suggesting the deviation of reality from the theoretical model is minor (see

Figure 2.4), but see the discussion for elaboration on these modifications.

Joint modelling of attributions for positive and negative events thus supported

three correlated cognitive style factors of internality, stability and globality, and two

uncorrelated affective biases on judgments of positive and negative events.

2.1.5 Replication final ASQ model

In order to test the replicability of the final model, an independent sample was next

collected. 232 undergraduates aged 17 – 21 years were recruited from a Chinese

university (97 male, 135 female) as participants in the replication study. All testing

procedures were identical, and no subjects from the replication study participated in

the previous study.

Replicability was tested by running the exact model constructed for Study one,

including the modifications required to raise that model to adequate fit. This model

showed an excellent fit between model and data (χ² (458) = 633.43, p <.001; CFI =

0.97; TLI = 0.96; RMSEA = 0.025; AIC = 984.61; BIC = 1695.13) (see Table 3.2).

The independent replication supported the structure found in previous study.

As the model fit well in both samples, we combined them in a final analysis to

maximize the precision of all estimated parameters. This also fit well (χ² (458) =

696.42, p <.001; CFI = 0.97; TLI = 0.96; RMSEA = 0.024; AIC = 1044.19; BIC

=1822.99) (see Figure 2.7, Table 2.2).



Figure 2.6: Well-fitting “3-cognitive styles + 2- explanatory styles” model of causal attributions for both positive and negative events.



Figure 2.7: Well-fitting joint model in the combined data set.



2.1.6 Schematic model of attributional style

Explanatory style models of optimism focus on three aspects of attributions about the

causes of positive and negative events: stability, pervasiveness, and internal-external

control. Within positive and negative event valences, these three aspects are

predicted to cluster forming explanatory style factors for each type of event, and

these in turn are predicted to correlate negatively, in line with attributional accounts

of depression. This structure was tested first in two studies including both positive

and negative events simultaneously, as well as controlling for non-independence of

responses within events.

Study one consisted of ASQ responses collected in 452 Chinese subjects. For

models containing only positive or only negative events, the proposed three

correlated-factor structure of explanatory style fit well. However, in joint models of

both positive and negative events, three strong correlated cognitive style factors

emerged, which applied to all events independent of valence. That is subjects who

described events as local or as stable in nature, tended to do so for both positive and

for negative events. In addition, two uncorrelated factors of attributions to positive

and to negative events emerged.

To validate this model, an independent sample of 232 subjects was collected

and the exact model from study one was confirmed as well fitting in this second

sample. The ASQ captures two major structures: A set of cognitive styles: tendencies

to process events as, for instance, internal or external in causation, and uncorrelated

factors of bias regarding positive and negative event bias.

Simply, here in two studies I tested the structure of attributions made

regarding the causes of positive and negative events (Abramson et al., 1978). Figure

2.8 shows in a schematic but quantitative form, the final conclusions emerging from

the joint analysis of the first study one and the replication study.



Figure 2.8: Final model reflecting results from combined data in two studies.

Analyses of single event valences revealed correlated globality, stability, and

internality factors as reported by Hewitt et al. (2004) and Higgins et al. (1999)

replicating in a non-Western sample the prior pattern and supporting the validity of

the scale in China. However the joint analyses revealed a very different outcome.

Attributional biases to positive events and to negative events emerged as

uncorrelated. Importantly, three valence-independent cognitive styles were required

to account for responding: global-local, stable-unstable, and internal-external. The

implications of these findings are discussed below.



Cognitive styles emerged as an important influence on responding: valence-

independent cognitive styles accounted for 85 percent of variance in the latent-factor

model. This suggests that subjects apply consistent cognitive styles independent of

event-valence, with personal tendencies to explain events as, for instance, global or

local independent of event valence: Subjects rating positive events as global tended

also to describe negative events in terms of global attributions, and likewise for the

other two styles. The cognitive styles correlated modestly, with coherent tendencies

to global-stable-internal vs local–unstable-external attributions.

It should be noted that several minor modifications were required to achieve

accepted levels of fit for this model. These mostly involved small item-item

correlations: this redundancy might allow a revised scale to be shortened. Eleven

changes were theoretically significant paths from cognitive style to attributions

outside the style: for instance from stability to globality of event 3. These indicate

that revision or deletion of some items may improve the diagnostic coherence and

utility of scales derived from well-fitting models of the ASQ.

Optimistic and pessimistic explanatory styles also emerged, with a pessimistic

explanatory style associated with beliefs that the causes of negative events are stable,

persuasive, and internal, and a positive bias for events being brief, affecting only one

aspect of life, and be externally caused (Forgeard & Seligman, 2012). Supporting

several empirical studies, optimistic and negative event were uncorrelated in the

present data (P.J. Corr & J.A. Gray, 1996; Peterson et al., 1982).

Based on these findings, attributions may be best viewed as reflecting large

differences in cognitive style (independent of event valence), and smaller

independent positive– and negative-event biases. Scoring and interpretation of the

ASQ should reflect this. Responses should be scored for cognitive style in addition to

optimistic or pessimistic explanatory bias. For most individuals, mixed attributional

styles should be expected: such as optimistic explanations for negative events and

pessimistic attributions for positive events.



2.2 Separating optimism and pessimism

2.2.1 Previous understanding of dispositional optimism

Psychometric structure of the LOT

As the most frequently used measure of dispositional optimism, the LOT or its

revised version, the LOT-R, has been applied widely in numerous studies. One

critical issue concerning the dimensionality of this instrument, is whether it measures

one dimension (optimism) or two dimensions (optimism and pessimism), is still not

quite clear. This dispute has been examined by a number of empirical studies with

controversial results demanding further investigation.

Theoretically, the basic conceptualization of dispositional optimism is formed

on the behavioural self-regulation model, addressing both goals approach and goals

avoidance (Carver & Scheier, 2001). Accordingly, expectancies should be involved

in both goal approach and gaol avoidance processes. Based on this framework,

dispositional optimism was originally assumed to be a bipolar dimension. Scheier

and Carver (1985) suggested that the LOT measured a one-dimensional bipolar

construct of dispositional optimism (n = 624). For the LOT-R, (Scheier et al., 1994)

proposed that “confirmatory factor analysis further indicated that the single-factor

solution was superior to a two-factor one” (n = 4,309).

At the same time, however, in a study with a sample of 889 male sailors in

the Navy (Marshall, Wortman, Kusulas, Hervig, & Vickers Jr, 1992), evidence

indicated that the positively and negatively phrased items in the measure split into

two factors. The factor of positively phrased items was named as “optimism”, and

the factor of negatively worded items was named as “pessimism”. This two-factor

model, which declared that optimism and pessimism represent two distinct traits, was

replicated in several later studies (Chang et al., 1997; L. Chang & McBrideChang,

1996; Creed et al., 2002; Roysamb & Strype, 2002). For example, in a sample of 347

undergraduates, Steed (2002) reported that the two-factor model was superior to the

one-factor model using a confirmatory factor analysis (CFA) approach. This two-

dimensional structure was replicated in an adolescent sample recently (Monzani et al.,

2014).



Though most studies support the two-factor theory, it is not clear whether this

two-dimensional model occurs through methodological bias or just reflects

substantive differences among items. To deal with this issue, Kubzansky, Kubzansky,

and Maselko (2004) reversed the framing of half of the items on each subscale, and

compared the method artefact model with the two-factor model. Their results

indicated that the bidimensional factor structure is consistent across all LOT versions

no matter how each item is framed. In addition, McPherson and Mohr (2005) tested

the potential effect of extremity of item wording on the LOT, and demonstrated that

item extremity had no influence on the bidimensional structure at all.

Though the dimensional dispute of dispositional optimism has been mainly

examined theoretically, there is at least one example in which the psychometric

structure of dispositional optimism was investigated by linking it to physical index.

Räikkönen and Matthews (2008) reported that while high pessimism predicted high

ambulatory blood pressure, low optimism had no effects on this physical index. It

indicated that dispositional optimism measured by the LOT may be not a bipolar

construct as originally assumed.

To summarize the findings to date, a two-factor structure is psychometrically

preferable to a one-dimension structure of total dispositional score (Suzanne C.

Segerstrom, Evans, & Eisenlohr-Moul, 2011). This bidimensional structure of

dispositional optimism was further supported in a large, age-heterogeneous sample

(46,133 participants aged from 18 to 103 years). Results indicated that the LOT-R is

bidimensional, consisting of an optimism factor and a pessimism factor. This two-

dimensional construct model was found to be stable across gender and age groups

(Herzberg, Glaesmer, & Hoyer, 2006).

Although different versions of the LOT or the LOT-R have been applied in a

variety of research on optimism during the past two decades, a majority of these

studies were conducted in Western samples. Consequently, it remains unclear for the

applicability of the concept and structure of dispositional optimism in Eastern

cultures. Sumi (2004) tested a measure of the Japanese translation of the LOT-R in

223 Japanese undergraduates. The original English version of the LOT or a Chinese



adaptation of the test has been conducted among Hong Kong Chinese (n = 620) and

Taiwanese (n = 1,119) (Cheng & Hamid, 1997; Li, 2012). Results of these studies

generally support the two-factor model that was found in most English-speaking

samples.

However, even in the few studies of optimism in non-English speaking

countries, controversy still exists. For instance, in a study of dispositional optimism

with Hong Kong Chinese (Lai, 1997), a modified Chinese version of the Life

Orientation Test was administered to one college student sample (n = 230) and an

adult sample (n = 173). The results indicated that the predictive power of the LOT

was owed to the optimism subscale. That is, the findings supported the

unidimensional view of the LOT. This evidence for the one-factor model was

replicated when the original English version of the LOT-R was applied in 248 Hong

Kong Chinese (Lai et al., 1998).

Until recently, studies of dispositional optimism have been rarely conducted

on Eastern cultures; and even fewer studies have been done with Mainland Chinese.

One of the exceptions was a study conducted by Lai (2000) in 404 Hong Kong

students and 328 Mainland Chinese students. A mixed scale of the LOT-R adaptation

and the Chinese version of the original LOT were completed by the participants.

CFA analysis indicated that while the bidimensional interpretation applied to the data

of the Mainland Chinese students, the Hong Kong sample showed a one-factor model.

To further apply the widespread measure of dispositional optimism in

Mainland China, it is necessary to examine the factor structure of the LOT-R in

Mainland Chinese samples. A proper examination of the applicability of

dispositional optimism to Chinese samples should apply translation of the LOT-R,

which is currently the most prevalent measure. By using the translated version of the

LOT-R, we attempted to provide results that are more generalizable to the scientific

literature and to Eastern samples.



Linking dispositional optimism to explanatory style

As two main approaches to conceptualizing and measuring optimism, dispositional

optimism and explanatory style have long been linked together and both have a wide

range of applicability in research with parallel findings with depression, well-being

and other related psychological constructs (Carver et al., 2010; Forgeard & Seligman,

2012).

Explanations for past events influence expectations for the future (Peterson &

Seligman, 1984). That is, if a person attributes past failures to causes that are stable,

he or she will expect more failures in the future, because the cause is likely to remain

for a long time. If the cause of a negative event is attributed to global factors, the

expectations tend to be that actions will not be under control even in many other

situations. In parallel, if the explanation for a negative event is explained by internal

factor, lower self-esteem tends to be displayed and passive expectation will be

produced. Scheier and Carver (1992) also pointed out that explanatory style and

dispositional optimism simultaneously rely on at least partly the same assumption,

which claims that differences in people’s expectations result in optimistic versus

pessimistic consequences.

In a study conducted by Metalsky et al. (1993), 114 college students subjects

were instructed to write down their expectations for their future performance on an

exam, after they completed the EASQ. The results indicated that among

undergraduates who received a low score, those who ascribed undesirable academic

performance to stable and global factors expected themselves to not achieve well in

the future. This result can be seen as evidence of potential influence of attributions

on expectations.

Though dispositional optimism and explanatory style are taken as

theoretically linked to each other, the results from empirical research exploring the

relationship between these two variables is inconsistent. Measures of generalized

expectancies (by the LOT) are only low or modestly associated with explanations for



negative events (by the ASQ) (Ahrens & Haaga, 1993; J. E. Gillham, Shatté, Reivich,

& Seligman, 2001; Peterson & Vaidya, 2001).

Generally speaking, correlations between the two constructs were positive but

varying between low and high across studies. Scheier and Carver (1992) reported

that correlations between the ASQ and the LOT are not very strong. Peterson and

Vaidya (2001) found a correlation of .20 between the ASQ and the LOT among a

sample of 155 college students. In one study conducted by Ahrens and Haaga (1993),

94 undergraduates completed several measures included the LOT and the ASQ, and

the correlation was reported as .30. In contrast, Hjelle, Belongia, and Nesser (1996)

reported a correlation of .41 between the LOT and the ASQ composite in a subject of

436 college students. J. E. Gillham, Tassoni, Engel, DeRubeis, and Seligman (1998)

reported a correlation of .63 and .41 between the LOT and the ASQ at two

assessment points. These correlations went up to .77 and .49 after being corrected for

attenuation respectively. Thus, correlations between the LOT and the ASQ ranged

from .20 to .77 across these studies.

Aims and hypothesis

Given the inconsistency in previous research, the current study examined two issues

regarding the nature of dispositional optimism. First, I wished to examine the utility

of a Chinese version of the LOT-R to measure dispositional optimism with a

Mainland Chinese sample. It is important to reach a resolution regarding the

psychometric structure of this popular measure of dispositional optimism before its

widespread application in Mainland China.

Based on previous findings mostly reporting a two-factor model of the LOT-

R, it is hypothesized that the two-factor model is superior to the one-factor model in

my study. Second, I set out to investigate the relationship of dispositional optimism

and explanatory style through correlational analysis. Based on previous findings, I

hypothesized that ASQ dimensions and LOT-R Optimism and LOT-R Pessimism

would be weakly correlated.



2.2.2 Two-factor structure of the LOT

A total of 684 participants, 452 from sample 1 and 232 of which from sample 2, were

included in this study (for detail of these two samples, see 1.5.4 of Chapter 1). There

were 230 males and 454 females. The mean age of the total sample was 19.93 years

(SD = 1.42).

Dispositional optimism was measured using the Chinese LOT-R (Lai & Yue,

2000).

Attributional style was assessed using the Chinese ASQ (Zhang, 2006).

Analysis Strategy

Structural equation modelling (SEM) was used to test potential mediating models

comprising the LOT-R using Amos 17.0 (Arbuckle, 2008). All analyses took

advantage of raw data supporting estimation of models using full information

maximum likelihood estimation. Descriptive statistics and correlational analyses

were obtained.



(RMSEA). For CFI and TLI, values > 0.95 were taken as indicating acceptable fit

(Hu & Bentler, 1999). For the RMSEA, values of < .05 indicated acceptable fit (C. Y.

Yu, 2002). Akaike Information Criterion (AIC) and Bayesian Information Criterion

(BIC) are reported to aid model comparison.



Modelling

I first tested the one-factor model; all six items were specified as indicators of a

single factor. The unidimensional model fit poorly with the data, with χ² (10, N = 684)

= 405.19, p < .001; CFI = .358; TFI = .306; RMSEA = .241; AIC = 427.193; BIC =

477.000.

I next turned to the two-factor model. Here the three positively worded items

were specified as indicators of the Dispositional Optimism factor (LOT-R Optimism),

and the three negatively worded items were specified as indicators of the

Dispositional Pessimism factor (LOT-R Pessimism). Compared with the one-factor

model, the two-factor model fit better with χ² (8, N = 684) = 26.525, p < .001; CFI

= .970; TFI = .944; RMSEA = .058; AIC = 52.525; BIC = 111.388 (See Figure 2.9).

The correlation between the Dispositional Optimism factor and the Dispositional

Pessimism factor was -.20 (p<.01). The factor loading ranged from .30 to .81 (See

Figure 2.9).

Thus, as previously reported by many studies conducted in the Westerners, a

two-factor model of dispositional optimism was supported in this Mainland Chinese

sample. That is, the LOT-R measures two negatively correlated and independent

constructs.



Figure 2.9: Standardized estimations for the two-factor model.

Descriptive statistics

The means, standard deviations and Cronbach’s alpha of the total samples on LOT-R

and ASQ are summarized in Table 2.3.

Correlational analysis

I next turned to examine correlations between dispositional optimism and

explanatory style. I hypothesized that LOT-R Optimism and LOT-R Pessimism and

ASQ dimensions would be weakly correlated. Table 2.4 shows the inter-correlations

among the variables of interest. Consistent with previous studies, the LOT-R

Optimism was positively correlated with the ASQ Total (r = .12, p < .01), but lower

than correlations between these two variables reported by earlier studies (r ranged

from .20 to .77). For individual dimensions, LOT-R Optimism was positively

correlated with ASQ Positive (r = .08, p < .05) and Stable Positive (r = .09, p < .05),

Dispositional

Optimism

LOT-R 10

.74

LOT-R 4 .69

LOT-R 1 .44

Dispositional

Pessimism

LOT-R 9

.81

LOT-R 7 .67

LOT-R 3 .30

-.20



and negatively correlated with Stable Negative (r = -.10, p < .05), but had no

significant correlation either with ASQ Negative or with any three dimensions of

negative events. No significant correlation was found between ASQ Pessimism and

any ASQ dimensions.


LOT-R Optimism 6.37 2.33 0.64

LOT-R Pessimism 4.89 2.04 0.61

ASQ Positive 15.22 1.88 0.83




ASQ Negative 12.93 1.83 0.78




ASQ Total 2.29 2.19 0.84

Table 2.3: Means, SDs and Cronbach’s Alpha for the ASQ and the LOT scales.

Note: Means for ASQ dimensions are on a scale ranging from 1 to 7; Means for the

LOT are on a scale ranging from 0 to 4 (n = 684).



Measures 1 2 3 4 5 6 7 8 9 10

1. LOT-R Optimism

2. LOT-R Pessimism -0.13**

3. ASQ Positive 0.08* 0.03

4. Internal Positive 0.05 0.05 0.77**

5. Stable Positive 0.09* 0.04 0.85** 0.54**

6. Global Positive 0.06 -0.02 0.81** 0.39** 0.52**

7. ASQ Negative -0.06 -0.05 0.30** 0.09* 0.25** 0.37**

8, Internal Negative 0.04 0.01 0.26** 0.27** 0.19** 0.17** 0.59**

9. Stable Negative -0.10* -0.04 0.16** -0.04 0.26** 0.16** 0.78** 0.20**

10.Global Negative -0.05 -0.07 0.26** 0.03 0.11** 0.46** 0.83** 0.28** 0.47**

11.ASQ Total 0.12** 0.07 0.61** 0.58** 0.52** 0.39** -0.58** -0.27** -0.52** -0.47**

Table 2.4: Correlations between measures.

* p < 0.05

** p < 0.01



2.2.3 What we should know about dispositional optimism

The primary goal of the current study was to address whether dispositional optimism

measured by LOT-R was compatible with a one-factor or two-factor model in a

Mainland Chinese sample. I found that the LOT-R was better interpreted as a

bidimensional construct, which includes dispositional optimism and dispositional

pessimism, than a unidimensional structure.

Originally, dispositional optimism was theoretically constructed on self-

regulation theory, which involves approaching and avoiding goals of behaviour, and

was then proposed to reflect a bipolar construct (Scheier & Carver, 1985). However,

many studies demonstrated that the two-factor structure may better explain the

psychometric structure of dispositional optimism (L. Chang & McBrideChang, 1996;

Kubzansky et al., 2004; Marshall et al., 1992; McPherson & Mohr, 2005; Roysamb &

Strype, 2002). The present study conducted in a Mainland Chinese sample supported

the proposal of a bidimensional construct. Though prior studies concerning the

psychometric structure of the LOT and LOT-R mainly support a two-factor model, it

does not mean that individuals should be distinctively categorized as optimists and

pessimists by a cut-off score. As noted in the study of Eichner, Kwon, and Marcus

(2014), optimism is a continuous variable.

The second aim of the present study was to examine the correlations between

dispositional optimism and explanatory style. Results indicated that dispositional

optimism was positively correlated with the composite attributional style, which is

consistent with most previous studies exploring the relationship between these two

constructs, although the correlation was lower than earlier studies. New findings were

reported for correlations between the LOT Optimism and individual dimensions of the

ASQ. Specifically, the results demonstrated that LOT-R optimism was positively

correlated with the stability dimension of positive events, and negatively correlated



with the stability dimension of negative events. This may reveal some interesting

points in understanding the relationship between dispositional optimism and

explanatory style. Regarding the fact that only a general correlation between the LOT

or LOT-R and the ASQ composite has been reported in most previous studies, results

in this study provide at least some further information to better understand the

relationship between dispositional optimism and explanatory style.

Furthermore, my study provided empirical evidence of the correlational patterns

between explanatory style and dispositional optimism in a non-Western sample. The

results were generally consistent with findings of previous research in Western

samples, in which explanatory style and dispositional optimism were reported to be

weakly correlated (Forgeard & Seligman, 2012).


Chapter 3: Optimism and personality 82

Chapter 3: Optimism and personality

A pessimist sees the difficulty in every opportunity. An optimist sees the opportunity in

every difficulty. – Winston Churchill

3.1 Is optimism a personality thing?

Personality, as one of the most traditional and widely developed psychological models,

has long been the focus of theorists and practitioners. There are at least three different

well-established personality systems – Eysenck’s three factor approach (Eysenck,

1965), the 16 personality factor system (Cattell, 1943), and the Five-Factor Model of

personality (FFM; McCrae & Costa, 1987) – that have been proposed and studied in

the last several decades. Among these three approaches, the FFM appears to have

attained a dominant position in both research and application.

The FFM proposes that there are five fundamental dimensions of personality that

are stable and consistent over time and across culture, namely Extraversion,

Agreeableness, Neuroticism (Emotional Stability), Conscientiousness, and Openness

to Experience (McCrae & Costa, 1987). The FFM is measured with the Revised NEO

Personality Inventory (NEO-PI-R; Costa & McCrae, 1992). Each of the five domains

of the NEO-PI-R is represented by six specific scales that measure facets of each

domain. For example, Neuroticism consists of Anxiety, Angry Hostility, Depression,

Self-Consciousness, Impulsiveness, and Vulnerability; Extraversion consists of

Warmth, Gregariousness, Assertiveness, Activity, Excitement-Seeking, and Positive

Emotions (Costa & McCrae, 1995).



Explanatory style and personality

Explanatory style has been proposed as a cognitive variable designed to investigate

the habitual causal explanations people provide for life events (Peterson & Seligman,

1984). Attributions are identified as thoughts and beliefs people hold for explaining

various life events, and this individual difference has been assessed largely through its

linkage to traditional personality traits, including almost all the main approaches in

personality. Previous studies have indicated that attributions for life events, especially

for negative events, provides understanding of the potential mechanism underlying

the nature of other personality dispositions (e.g. Haugen & Lund, 1998).

Though both explanatory style and FFM have been taken as important to

understanding personality, very few studies have been done to explore the relationship

between these two constructs. In those studies, attributional style for negative events

has been found to be negatively correlated with Conscientiousness. For example, in a

study which investigated substance use in college students, Musgrave-Marquart et al.

(1997) reported that attributions for academic failure was modestly correlated with

Conscientiousness (r = -.18) but none of the other FFM dimensions. Similarly,

Poropat (2002) reported that ASQ Negative was negatively correlated with

Conscientiousness (r = -.16). Correlations between ASQ Positive, ASQ Total, and

FFM dimensions have also been reported in this study. ASQ Positive was found to be

positively correlated with Emotional Stability (r = .18) but not significantly associated

with other FFM dimensions. By contrast, ASQ Total has been reported to correlate

significantly with four FFM dimensions (Extraversion, r = .22; Agreeableness, r = .16;

Conscientiousness, r = .20; Emotion Stability, r = .22).

In addition to FFM, correlations between ASQ dimensions and other personality

frameworks have been investigated. For example, Haugen and Lund (1998) reported

that attributions for positive and negative events correlated differently with self-



esteem, motive, self-efficacy, and defensiveness. In a group of Chinese college

students, Wang and Zhang (2005) reported correlations between the ASQ and the

Sixteen Personality Factor Questionnaire (16-PF). It revealed that individuals with a

pessimistic explanatory style were also characterized by high sensitivity, high

insecurity, high tension, and high anxiety.

In their study of ASQ validation, P.J. Corr and J.A. Gray (1996) examined ASQ

correlations with several personality traits from the Eysenck Personality

Questionnaire (EPQ) and the State-Trait Anxiety Inventory (STAI). Attributions for

positive events correlated positively with Extraversion within the occupational sample

of salespersons but did not correlate with any of the EPQ variables among a group of

volunteers. Attributions for negative events was correlated with all EPQ variables,

suggesting a trend of general dysphoria, e.g. high Neuroticism, high psychoticism,

and low Extraversion, which was consistent with a general understanding of the

relationship between negative attributional style and the FFM. On the other hand,

anxiety measured using the STAI correlated positively with ASQ negative events

scores and negatively with the ASQ positive events scores.

Studies examining the relationship between explanatory style and personality

have often been intertwined with the investigation of potential gender differences in

attributional style. For instance, Rim (1991) reported that for the dimension of

stability, men scoring low on Neuroticism rated higher on positive than negative

events, while for the global factor, those scoring high on Neuroticism rated higher on

positive than on negative events. Women have different patterns. For all attributional

styles, women who scored low on Neuroticism had higher scores on positive events

than on negative events. Regarding Extraversion, both men and women with low

scores got higher rates on positive than on negative events for the internal factor only.



Gender differences were also reported in a later study. Poropat (2002)

investigated the relationship between explanatory style and the FFM in a group of

college students, and discovered that the correlational patterns were different for men

and women. Specifically, optimistic explanatory style was positively related to

Agreeableness for both men and women, but was positively related to Extraversion

only for men, and was negatively related to Neuroticism only for women.

Gender differences have also been reported in studies examining the link between

explanatory style and other basic personality variables in addition to the FFM. For

example, using the California Psychological Inventory (CPI) as a personality

measurement, Bunce and Peterson (1997) reported that women’s optimistic

explanatory style negatively correlated with well-being and good impression. For men,

different patterns emerged. Sociability negatively correlated with optimistic

explanatory style. Though the mechanism underlying the gender differences in the

attributional style-personality relationship is still not quite clear, these studies indicate

that they are manifested differently between men and women.

Based on the prior studies mentioned above, it appears that there are no consistent

pattered correlations between explanatory style and FFM variables and other

personality frameworks. This lack of research called for the necessity of studies

comparing these two important variables.

Dispositional optimism and FFM

Dispositional optimism is regarded as a relatively stable individual personality trait

(Carver et al., 2010). Associations between dispositional optimism and the FFM have

been found in many studies. Dispositional optimism is mainly manifested in

Neuroticism and Extraversion, especially the former. For example, Williams (1992)

reported that the LOT correlates positively with Extraversion (r = .25), and is also

correlated negatively, but more strongly, with Neuroticism (r = -.58). This study was



conducted with 223 university students. Also, in a sample of 113 older women,

Boland and Cappeliez (1997) linked optimism to low Neuroticism (r = -.66).

Significant correlations between dispositional optimism and other FFM

dimensions have been reported. For example, Suzanne C Segerstrom, Castañeda, and

Spencer (2003) reported strong positive correlations between LOT-R scores and

Conscientiousness (r = .31), in addition to typical correlations of dispositional

optimism with Extraversion (r = .60) and Emotional Stability (r = -.46). Furthermore,

Agreeableness was found to be positively correlated with dispositional optimism in

Ebert, Tucker, and Roth (2002)’s study (r = .35). The relationship between

dispositional optimism and the FFM was expanded to Openness as well. Lounsbury,

Saudargas, and Gibson (2004) reported positive correlations between dispositional

optimism and all five FFM dimensions: Extraversion (r = .27), Conscientiousness (r

= .23), Agreeableness (r = .29), Emotional Stability (r = .60), and Openness (r = .30).

Similarly, in a larger-sample study (N = 4,332), Sharpe, Martin, and Roth (2011)

reported that dispositional optimism (measured by three different questionnaires) was

significantly correlated with all five FFM factors (assessed by five different measures).

For Extraversion, raverage = .44; for Neuroticism, raverage = -.56; for Openness, raverage

= .21; for Agreeableness, raverage =.39; for Conscientiousness, raverage = .38.

One of the unresolved debates about dispositional optimism is whether it is a

continuous bipolar variable or a two-dimensional variable. Implied in measurement,

there has long been an ambiguity in confirming the psychometric structure of the LOT.

A few studies have tried to resolve this debate by linking dispositional optimism to

some traditional and well-established personality constructs, such as the FFM. In

these studies, the FFM or other fundamental personality traits have been used as

external criteria to examine the psychometric structure and personality essence of

dispositional optimism. For instance, a two-dimension structure of the LOT was



supported in Marshall et al. (1992)’s study in a sample of 889 male navy interns. This

study discovered that LOT Optimism correlated more strongly with Extraversion than

did LOT Pessimism, and LOT Pessimism correlated more strongly with Neuroticism

than did LOT Optimism, showing that the LOT is related to both these domains of

personality. However, the patterns revealed in Marshall et al.’s research were greatly

reduced after item valence was controlled for in a recent study with a larger sample

size (n = 1,016) (Kam & Meyer, 2012).

Aims of the current study

The present study set out to accomplish four main goals.

First, correlational analysis of ASQ measures, LOT-R variables and FFM factors

were calculated and these analyses were expanded to specific facets of FFM

dimensions in order to get a better understanding of the relationship between

explanatory style and dispositional optimism, and to provide extra information

concerning the relationship between optimism and the FFM. Based on previous

research findings already discussed, LOT-R Optimism was hypothesized to be

negatively related to Neuroticism, and positively correlated with Extraversion,

Agreeableness, Openness, and Conscientiousness. Conversely, LOT-R Pessimism

was hypothesized to be positively related to Neuroticism, and negatively correlated

with the other four FFM factors.

For the ASQ measures, ASQ Positive was hypothesized to be negatively related

to Neuroticism and positively correlated to Extraversion. ASQ Negative was

hypothesized to be negatively related to Conscientiousness. Other potential

correlations between ASQ variables and FFM factors, such as correlations between

ASQ Negative and Extraversion, have not been reported previously. Based on past

correlational analysis between FFM factors, ASQ Positive was hypothesized to be



positively related to Openness, Agreeableness, and Conscientiousness. ASQ Negative

was hypothesized to be positive correlated with Neuroticism, and negatively related to

Extraversion, Openness, and Agreeableness.

For specific facets of FFM factors, Depression (one of the six facets of

Neuroticism) was hypothesized to be positively related to ASQ Negative and LOT-R

Pessimism, and to be negatively correlated with ASQ Positive and LOT-R Optimism.

Other potential correlations between ASQ, LOT-R variables, and FFM facets are

unpredictable since no findings have been reported as to my knowledge.

Second we wished to explore gender difference in levels of explanatory style

within the background of FFM as suggested by Poropat (2002). Examination of

gender differences was extended to the relationship of dispositional optimism and

FFM variables. This study set out to compare the ASQ, the LOT, and the FFM among

men and women collectively as well as among men and women separately.

Third, since previous studies have suggested the FFM is a reliable external

criterion for examining the psychometric structure of dispositional optimism, the next

aim of this study was to test the associations between the FFM and dispositional

optimism/pessimism. In addition to correlational analyses, a model using SEM was

examined (see Figure 3.1). For this model, we hypothesized that all FFM dimensions

are correlated with each other; LOT-R Optimism and LOT-R Pessimism will be

predicted by FFM factors, especially Neuroticism and Extraversion; and LOT-R

Optimism and LOT-R Pessimism are distinctive but negatively correlated factors.



Figure 3.6: Proposal for an initial model with hypothesized relationship between

LOT-R and FFM.

Finally, I set out to examine the relationship between attributional style and FFM

with a SEM model (see Figure 3.2). In my earlier MTMM analysis of the ASQ, joint

modelling of attributions supported three correlated cognitive style factors of

internality, stability and globality, and two uncorrelated affective biases on judgments

of positive and negative events. Accordingly, in this model, it was hypothesized that

Internal Positive and Internal Negative are positively correlated, as are the other two

cognitive style factors (Stability and Globality). All FFM dimensions are correlated

LOT-R Optimism

LOT-R 10 LOT-R 4 LOT-R 1

LOT-R Pessimism

LOT-R 9 LOT-R 7 LOT-R 3

Neuroticism Extraversion Openness Agreeableness Conscientiousness



with each other. ASQ Positive and ASQ Negative will be predicted by FFM

dimensions. Specifically, Neuroticism and Extraversion were expected to be

predictors of attributional style.

Figure 3.2: Proposal for an initial model with hypothesized relationship between ASQ

and FFM.

Neuroticism Extraversio

n

Openness Agreeableness Conscientiousnes

s

ASQ Positive ASQ Negative

Global Stable Internal Global Stable Internal



3.2 Methods

Participants

A total of 452 participants (sample 1) were included in the current study (for detail of

this sample, see 1.5.4 of Chapter 1).

Materials

Dispositional optimism was measured using a Chinese version of the Life Orientation

Test-Revised (Lai & Yue, 2000). Subjects were scored for LOT-R Optimism and

LOT-R Pessimism scores. Cronbach’sαfor LOT-R Optimism, .76; and, for LOT-R

Pessimism, .82.

Attributional style was assessed using the Chinese ASQ (Zhang, 2006).

Composite attributional styles were calculated separately for positive and negative

events separately. Reliabilities (Cronbach’s α) were acceptable .84 for the total and,

for positive events, .84; for negative events, .77.

Though the NEO-PI-R is a well-established, psychometrically sound instrument

that covers a full range of the Big Five personality traits, it has rarely been used in

prior research partly due to its time-consuming length. The FFM was measured, in the

present study, by a Chinese version of the NEO-PI-R (Yang et al., 1999). The internal

consistency of the personality total from the NEO-PI-R was .83 in this sample.

Reliabilities (Cronbach’s α) were acceptable for five individual sub-scales (.89 for

Neuroticism; .83 for Extraversion; .76 for Openness; .75 for Agreeableness; and .88

for Conscientiousness).



Analysis Strategy

Structural equation modelling (SEM) was used to test potential models constructing

LOT-R and NEO-PI-R using Amos 17.0 (Arbuckle, 2008). All analyses took

advantage of raw data supporting the estimation of models using full information

maximum likelihood estimation. Descriptive statistics and correlational analyses were

obtained.


Tucker-Lewis index (TLI), and the Root Mean Square Error of Approximation

(RMSEA). For CFI and TLI, values > 0.95 were taken as indicating acceptable fit (Hu

& Bentler, 1999). For the RMSEA, values of < .05 indicated acceptable fit (C. Y. Yu,

2002). Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC)

are reported to aid model comparison.

3.3 Results


Table 3.1 demonstrates descriptive statistics andαreliability coefficients for the ASQ

and the LOT-R scales. The ASQ reliabilities reported in Table 3.1 are similar to those

reported by Peterson et al. (1982) and Poropat (2002).




LOT-R optimism 8.27 1.84 0.76

LOT-R pessimism 3.85 1.99 0.82

ASQ Negative 12.90 1.78 0.84




ASQ Positive 15.28 1.91 0.77




ASQ Total 2.38 2.17 0.84

Table 3.1: Means, standard deviations and Cronbach’s alpha for ASQ and LOT-R

scales.

Descriptive statistics and Cronbach’s alpha for NEO-PI-R scales are reported in Table

3.2.




Neuroticism 94.10 18.81 0.89

Anxiety 16.87 4.32 0.66

Angry Hostility 13.58 4.13 0.62

Depression 15.29 4.40 0.68

Self-consciousness 17.68 4.17 0.63

Impulsiveness 15.58 3.57 0.50

Vulnerability 15.08 3.98 0.69

Extraversion 106.15 15.39 0.83

Warmth 20.90 4.25 0.72

Gregariousness 17.16 4.24 0.65

Assertiveness 15.08 3.64 0.60

Activity 16.01 3.33 0.42

Excitement-seeking 15.81 3.49 0.34

Positive Emotions 21.19 4.60 0.75

Openness 109.44 13.47 0.76

Fantasy 17.74 3.99 0.61

Aesthetics 20.01 4.13 0.62

Feelings 20.46 3.62 0.57

Actions 14.97 3.19 0.41

Ideas 17.31 4.71 0.74

Value 18.95 2.89 0.23

Agreeableness 112.52 12.31 0.75

Trust 20.05 3.65 0.63

Straightforwardness 17.73 3.83 0.54

Altruism 21.32 3.66 0.64

Compliance 18.10 3.34 0.37

Modesty 15.22 2.99 0.42

Tender-Mindedness 20.09 3.46 0.46

Conscientiousness 111.14 17.15 0.88

Competence 18.62 3.42 0.53

Order 16.97 3.90 0.56

Dutifulness 21.08 3.68 0.58

Achievement Striving 18.18 4.23 0.68

Self-Discipline 17.55 3.66 0.62

Deliberation 18.74 4.10 0.67

Table 3.2: Means, standard deviations and Cronbach’s alpha for NEO-PI-R scales.

Correlational analyses

I first tested correlations between the ASQ, the LOT-R and the five NEO-PI-R scales

for the entire sample (see Table 3.3). Both LOT-R Optimism and ASQ Total have



significantly negative correlations with Neuroticism, and significantly positive

correlations with Extraversion, which is consistent with prior studies (e.g. Poropat,

2002; Sharpe et al., 2011). Both LOT-R Optimism and ASQ Total are significantly

correlated with Openness, Agreeableness, and Conscientiousness for the entire sample.

LOT-R Pessimism is positively correlated with Neuroticism and negatively correlated

with Extraversion, Openness, and Conscientiousness, but not significantly correlated

with Agreeableness.

As expected, ASQ Positive and ASQ Negative have different correlational

patterns with the FFM. ASQ Negative is positively correlated with Neuroticism, and

is negatively correlated with Extraversion and Conscientiousness, while ASQ Positive

is positively related to four of the five NEO-PI-R dimensions (see Table 3.2).

Measures Neuroticism Extraversion Openness Agreeableness Conscientiousness

LOT-R Optimism -0.32** 0.40** 0.21** 0.22** 0.27**

LOT-R Pessimism 0.23** -0.26** -0.14** -0.09 -0.25**

ASQ Negative 0.31** -0.20** -0.04 -0.09 -0.23**

ASQ Positive -0.07 0.15** 0.22** 0.11* 0.19**

ASQ total -0.32** 0.30** 0.23** 0.17** 0.36**

Table 3.3: Correlations of LOT, ASQ and NEO-PI-R for the entire sample.

*P<0.05. **P<0.01.



To compare potential gender differences between the relationships of the LOT-R,

ASQ, and FFM, these correlations are demonstrated separately for men and women in

Table 3.4 and Table 3.5.

I first compared patterns of associations between men and the entire group. As

shown in Table 3.4, correlational patterns between the LOT-R, ASQ and NEO-PI-R

are quite similar for men and for the entire sample but still show differences. The

significant correlation between LOT-R Optimism and Openness for the entire sample

is absent for men. Similar patterns emerge for correlations between LOT-R Pessimism

and Openness. However, LOT-R Pessimism is negatively correlated with

Agreeableness for men while this correlation is absent for the entire sample.


LOT-R Optimism -0.30** 0.35** 0.14 0.28** 0.41**

LOT-R Pessimism 0.18* -0.26** -0.13 -0.22* -0.23**

ASQ Negative 0.38** -0.21* 0.01 -0.04 -0.18*

ASQ Positive -0.01 0.17* 0.23** 0.19* 0.18*

ASQ Total -0.35** 0.34** 0.20* 0.20** 0.32**

Table 3.4: Correlations of LOT-R, ASQ and NEO-PI-R scales for men.

*P<0.05. **P<0.01.

Then correlational patterns of these variables between men and women were

compared. Slight differences emerge (see Table 3.5). There is a positive correlation

between LOT-R Optimism and Openness for women, which is absent among men.

This is also the case for the negative correlation between LOT-R Pessimism and

Openness. However, the negative correlation between ASQ Negative and

Agreeableness for men is absent for women. Also, while ASQ Positive is positively

correlated with Agreeableness for men, it is absent among women.



In addition to the correlational analysis between the ASQ, LOT-R, and the five

main domains measured by the NEO-PI-R, correlations between the ASQ, LOT-R,

and all NEO-PI-R facets for each domain for the entire sample were also calculated

(see Table 3.6 to Table 3.10). These correlational analyses were aimed to examine the

relationships among dispositional optimism, explanatory style, and specific

personality facets described by the NEO-PI-R.


LOT-R Optimism -0.34** 0.42** 0.23** 0.19** 0.20**

LOT-R Pessimism 0.26** -0.26** -0.14* -0.02 -0.27**

ASQ Negative 0.28** -0.20** -0.06 -0.11* -0.26**

ASQ Positive -0.10 0.14** 0.22** 0.08 0.20**

ASQ Total -0.31** 0.28** 0.24** 0.16** 0.38**

Table 3.5: Correlations of LOT-R, ASQ and NEO-PI-R scales for women.

*P<0.05. **P<0.01.

As shown in Table 3.6, LOT-R Optimism is negatively correlated with all six

facets of Neuroticism, and LOT-R Pessimism is positively correlated with all

Neuroticism facets. For attributional style, Hopelessness (stability + globality of

negative events) was significantly positively associated with all six facets of

Neuroticism, including Depression, which is consistent with the hopelessness theory

of depression (Abramson et al., 1989) and findings reported by Peterson and Vaidya

(2001). Here ASQ Negative is significantly associated with all six facets in addition to

Neuroticism but ASQ Positive is not, which supports the lack of a correlation between

ASQ Positive and ASQ Negative (Peterson et al., 1982).



Table 3.7 displays the correlations among LOT-R, ASQ, and Extraversion and its

six facets, namely Warmth, Gregariousness, Assertiveness, Activity, Excitement-

seeking, and Positive emotions. Prior research found significant correlations between

optimism and positive affect (Ahrens & Haaga, 1993; Daukantaite & Zukauskiene,

2012; Scheier & Carver, 1992), which was supported here (see Table 3.7).

Specifically, LOT-R Optimism and ASQ Positive are positively correlated with

Positive emotions, while LOT-R Pessimism and ASQ Negative are negatively related

to Positive emotions.

Table 3.8 provides results of correlational analyses of the LOT-R, the ASQ scales,

and all facets of the Openness factor. As shown in Table 3.8, both LOT-R Optimism

and ASQ Positive are positively correlated with four of the six facets of Openness,

including Aesthetics, Feelings, Ideas, and Value. On the other hand, while LOT-R

Pessimism is negatively associated with Feelings and Value, ASQ Negative shows no

significant correlations with these two facets but is negatively correlated with Actions

and is positively correlated with Fantasy.

Correlations between dispositional optimism, explanatory style, and six facets of

Agreeableness are reported in Table 3.9. Here LOT-R Pessimism and ASQ Negative

demonstrate similar patterns of correlation. Though these two scales are not

significantly associated with Agreeableness as a whole, both are negatively correlated

with Trust, Altruism, and Modesty. For LOT-R Optimism and ASQ Positive, similar

correlational patterns appear. Both scales are significantly associated with Trust,

Altruism, Modesty, and Tender-Mindedness in addition to their positive correlation

with Agreeableness.

Table 3.10 presents correlations among the LOT-R, ASQ scales, and six facets of

Conscientiousness. Here the correlational patterns are quite similar. Specifically, both

LOT-R Optimism and ASQ Positive are positively correlated with Conscientiousness

as a whole and all six facets of Conscientiousness. On the other hand, LOT-R



Pessimism and ASQ Negative demonstrate negative correlations with both

Conscientiousness and all the six facets.



Measures Neuroticism Anxiety Angry Hostility Depression Self-consciousness Impulsiveness Vulnerability

LOT-R Optimism -0.32** -0.24** -0.20** -0.32** -0.26** -0.14** -0.29**

LOT-R Pessimism 0.23** 0.18** 0.17** 0.27** 0.14** 0.15** 0.12*

ASQ Negative 0.31** 0.21** 0.24** 0.28** 0.25** 0.17** 0.29**

Internal Negative 0.12** 0.05 0.09 0.10* 0.11** 0.08 0.14**

Stable Negative 0.26** 0.17** 0.21** 0.24** 0.21** 0.15** 0.23**

Global Negative 0.28** 0.22** 0.20** 0.25** 0.21** 0.14** 0.25**

Hopelessness 0.32** 0.23** 0.24** 0.28** 0.25** 0.17** 0.29**

ASQ Positive -0.07 -0.06 -0.10* -0.07 0.02 -0.05 -0.07

Internal Positive -0.17** -0.16** -0.13** -0.15** -0.06 -0.11* -0.16**

Stable Positive -0.07 -0.07 -0.09 -0.05 0.01 -0.04 -0.08

Global Positive 0.04 0.07 -0.03 0.02 0.08 0.02 0.04

Hopefulness -0.01 0.00 -0.07 -0.02 0.05 -0.01 -0.02

ASQ Total -0.32** -0.22** -0.28** -0.29** -0.19** -0.18** -0.30**

Table 3.6: Correlations of LOT-R, ASQ and Neuroticism and its six facets for the entire sample.

*P<0.05. **P<0.01.



Measures Extraversion Warmth Gregariousness Assertiveness Activity Excitement-seeking Positive Emotions

LOT-R Optimism 0.40** 0.35** 0.19** 0.28** 0.23** 0.13** 0.33**

LOT-R Pessimism -0.26** -0.23** -0.17** -0.16** -0.14** -0.06 -0.24**

ASQ Negative -0.20** -0.15** -0.08 -0.21** -0.17** -0.05 -0.15**

Internal Negative -0.11* -0.08 -0.08 -0.07 -0.08 -0.05 -0.06

Stable Negative -0.20** -0.16** -0.05 -0.22** -0.17** -0.02 -0.17**

Global Negative -0.13** -0.09 -0.05 -0.14** -0.11* -0.04 -0.10*

Hopelessness -0.20** -0.14** -0.06 -0.21** -0.17** -0.04 -0.16**

ASQ Positive 0.15** 0.19** 0.05 0.07 0.06 0.07 0.12**

Internal Positive 0.13** 0.13** 0.01 0.12* 0.11* 0.04 0.11*

Stable Positive 0.15** 0.18** 0.06 0.11* 0.02 0.09 0.11*

Global Positive 0.08 0.14** 0.05 -0.03 0.02 0.05 0.07

Hopefulness 0.13** 0.19** 0.06 0.04 0.02 0.08 0.11*

ASQ Total 0.30** 0.29** 0.11* 0.23** 0.19** 0.10* 0.24**

Table 3.7: Correlations of LOT-R, ASQ and Extraversion and its six facets for the entire sample.

*P<0.05. **P<0.01.



Measures Openness Fantasy Aesthetics Feelings Actions Ideas Value

LOT-R Optimism 0.21** 0.05 0.13** 0.17** 0.04 0.19** 0.13**

LOT-R Pessimism -0.14** -0.01 -0.08 -0.15** -0.08 -0.08 -0.14**

ASQ Negative -0.04 0.12* -0.08 -0.02 -0.09* -0.06 -0.01

Internal Negative -0.06 0.06 -0.06 -0.06 -0.11* -0.03 -0.04

Stable Negative -0.05 0.09 -0.10* -0.05 -0.02 -0.10* 0.04

Global Negative 0.02 0.10* -0.02 0.05 -0.08 0.01 -0.02

Hopelessness -0.02 0.11* -0.07 0.01 -0.06 -0.06 0.01

ASQ Positive 0.22** 0.08 0.17** 0.24** 0.03 0.13** 0.15**

Internal Positive 0.17** 0.04 0.11* 0.16** 0.07 0.14** 0.07

Stable Positive 0.18** 0.05 0.11* 0.18** 0.06 0.09* 0.17**

Global Positive 0.19** 0.11* 0.18** 0.24** -0.04 0.08 0.12*

Hopefulness 0.22** 0.09 0.17** 0.24** 0.01 0.10* 0.16**

ASQ Total 0.23** -0.02 0.21** 0.23** 0.11* 0.16** 0.15**

Table 3.8: Correlations of LOT-R, ASQ and Openness and its six facets for the entire sample.

*P<0.05. **P<0.01.



Measures Agreeableness Trust Straightforwardness Altruism Compliance Modesty Tender-Mindedness

LOT-R Optimism 0.22** 0.37** 0.05 0.32** 0.05 -0.29** 0.21**

LOT-R Pessimism -0.09 -0.16** -0.02 -0.18** -0.05 0.24** -0.10*

ASQ Negative -0.09 -0.14** -0.05 -0.15** -0.04 0.15** -0.03

Internal Negative -0.03 -0.03 -0.02 -0.10* -0.05 0.08 0.01

Stable Negative -0.13** -0.16** -0.04 -0.16** -0.04 0.10* -0.11*

Global Negative -0.02 -0.11* -0.04 -0.08 -0.01 0.15** 0.04

Hopelessness -0.08 -0.16** -0.05 -0.14** -0.03 0.14** -0.04

ASQ Positive 0.11* 0.24** -0.09* 0.22** 0.06 -0.24** 0.17**

Internal Positive 0.06 0.22** -0.05 0.16** -0.01 -0.26** 0.08

Stable Positive 0.08 0.22** -.010* 0.19** 0.06 -0.25** 0.12*

Global Positive 0.13** 0.16** -0.08 0.18** 0.08 -0.10* 0.21**

Hopefulness 0.12** 0.22** -0.11* 0.21** 0.08 -0.20** 0.19**

ASQ Total 0.17** 0.33** -0.05 0.31** 0.09 -0.34** 0.18**

Table 3.9: Correlations of LOT-R, ASQ and Agreeableness and its six facets for the entire sample.

*P<0.05. **P<0.01.



Measures Conscientiousness Competence Order Dutifulness Achievement Striving Self-Discipline Deliberation

LOT-R Optimism 0.27** 0.33** 0.10* 0.18** 0.22** 0.25** 0.15**

LOT-R Pessimism -0.25** -0.22** -0.15** -0.10* -0.21** -0.25** -0.19**

ASQ Negative -0.23** -0.19** -0.16** -0.11* -0.15** -0.26** -0.17**

Internal Negative -0.15** -0.15** -0.13** -0.04 -0.12* -0.13** -0.12*

Stable Negative -0.19** -0.17** -0.12* -0.12* -0.10* -0.21** -0.14**

Global Negative -0.16** -0.11* -0.11* -0.07 -0.12* -0.21** -0.12*

Hopelessness -0.21** -0.16** -0.13** -0.11* -0.13** -0.25** -0.15**

ASQ Positive 0.19** 0.18** 0.11* 0.14** 0.16** 0.09* 0.16**

Internal Positive 0.21** 0.18** 0.14** 0.15** 0.18** 0.12** 0.18**

Stable Positive 0.17** 0.19** 0.11* 0.13** 0.14** 0.08 0.10*

Global Positive 0.10* 0.09 0.03 0.09 0.09 0.04 0.13**

Hopefulness 0.15** 0.16** 0.08 0.12** 0.13** 0.07 0.13**

ASQ Total 0.36** 0.32** 0.23** 0.22** 0.27** 0.29** 0.28**

Table 3.10: Correlations of LOT-R, ASQ and Conscientiousness and its six facets for the entire sample.

*P<0.05. **P<0.01.



SEM modelling for LOT-R and FFM

The proposed model (see Figure 3.1) between dispositional optimism and FFM was

tested (as shown in Figure 3.3).

Figure 3.3: Standardized estimations for the initial model for LOT-R and FFM.

Standardized estimates of the original model are shown in Figure 3.3. Chi-

square for the initial model was significant (χ² (28) = 86.74, p < .001). For the initial

LOT-R Optimism

LOT-R 10

.48

LOT-R 4

.67

LOT-R 1

.24

LOT-R Pessimism

LOT-R 9

.71

LOT-R 7

.73

LOR-R 3

.26


-.40 .45 .23 .33

-.21

-.16

-.52

.27

.42

.35

.03

-.26 -.02 .01

-.15

-.19

-.05

.11 .04 .35 -.41



base model, other index values were obtained as: CFI = 0.933; TLI = 0.869; RMSEA

= 0.068; AIC = 162.736; BIC = 319.056. Although CFI or TLI values may be

considered acceptable, modifications were suggested and made to the original model

to obtain a better fit according to the results. These modifications include three

relationships between the residual variances of measured variables, including a

relationship between the residual variance of Neuroticism and the first item of LOT-

R. The new paths all had loadings of .16 or below, suggesting that deviation from the

theoretical model is minor (see Figure 2.4). These modifications significantly

improved model fit, and the resultant model fit reasonably well (χ² (25) = 41.95, p

=.018; CFI = 0.981; TLI = 0.957; RMSEA = 0.039; AIC = 123.945; BIC = 292.606).

Figure 3.4: Standardized estimations for the modified model for LOT-R and FFM.

LOT-R Optimism

LOT-R 10

.47

LOT-R 4

.67

LOT-R 1

.32

LOT-R Pessimism

LOT-R 9

.71

LOT-R 7

.74

LOT-R 3

.24


-.41 .45 .23 .33

-.21

-.18

-.51

.27

.42

.35

-.45

.01

.34 -.26 .04 -.02 .11

.01

-.07

-.16

-.18

.16

-.13

.15



As shown in Figure 3.4, modelling analysis supports the initial model

proposed in Figure 3.1. In this model, Extraversion predicts both LOT-R Optimism

and LOT-R Pessimism with coefficients of .34 and -.26, respectively. Neuroticism

predicts only LOT-R Optimism (standardized coefficient = -.45). All FFM

dimensions are correlated with each other, with Neuroticism negatively correlated

with the four other FFM factors.

Multi-group SEM for testing gender differences of the model LOT-R and FFM

To formally test the potential gender differences of the model for LOT-R and FFM

(see Figure 3.4), multi-group SEM was conducted. See details in Chapter 3.3. I first

tested this model in the male group. Fit measures for this model indicated excellent

fit between model and data (χ2 (24) = 24.60, p < 0.5; CFI = 0.99, TLI = 0.99,

RMSEA = .014). Then, this model was tested in the female group. This model

showed a good fit between model and data (χ2 (24) = 34.35, p < 0.1; CFI = 0.98, TLI

= 0.96, RMSEA = .037).

Finally, multi-group SEM was conducted to test gender differences of this

model. In addition to unconstrained base model, Measurement weights, Structural

covariances, and Measurement residuals were used as constrained conditions in multi

group analysis. The fit statistics for baseline comparisons of all models tested are laid

out in Table 3.11. Table 3.11 shows that the unconstrained model fits best for the

data. Three constrained models have similar fits as the unconstrained model. Thus,

this model is compatible in both male and female in this sample.



Model NFI

Delta1

RFI

rho1

IFI

Delta2

TLI

rho2 CFI △CFI

Unconstrained .942 .867 .989 .972 .988

Measurement weights .908 .837 .967 .939 .965 -.023

Structural covariances .908 .845 .970 .947 .969 -.019

Measurement residuals .867 .837 .951 .939 .950 -.038

Table 3.11: Baseline comparisons for tested models between LOT-R and FFM.

SEM modelling for ASQ and FFM

The proposed model (see Figure 3.2) between dispositional optimism and FFM was

tested (as shown in Figure 3.5).



Figure 3.5: Standardized estimations for the initial ASQ-FFM model.

Standardized estimates of the original model are shown in Figure 3.5. Chi-

square for the initial model was significant (χ² (26) = 88.75, p < .001). For the initial

base model, other index values were obtained as: CFI = 0.950; TLI = 0.893; RMSEA

= 0.073; AIC = 168.754; BIC = 333.301. Although CFI or GFI values may be

considered acceptable, modifications were suggested and made to the original model



Global

.56

Stable

.80

Internal

.70

Global

.61

Stable

.59

Internal

.33

.58 .45 .30

.04 .32

.05 -.15 .18 .12

.01 -.05

.17

-.09

-.40 .45 .23 .33

-.21 .27 .35

-.16 .42

-.52



to obtain a better fit according to the results. These modifications include four

relationships between the residual variances of measured variables, for instance a

relationship between the residual variance of Neuroticism and ASQ Internal Positive.

The new paths all had loadings of .23 or below, suggesting the deviation from the

theoretical model is minor (see Figure 2.4). These modifications significantly

improved model fit, and the resultant model fit reasonably well (χ² (22) = 37.17, p

=.023; CFI = 0.988; TLI = 0.969; RMSEA = 0.039; AIC = 125.168; BIC = 306.170).

Figure 3.6: Standardized estimations for the modified ASQ-FFM model.



Global

.62

Stable

.81

Internal

.73

Global

.68

Stable

.62

Internal

.33

.64 .60 .18

.11 .31

.06 -.12 .17 .11 .01

-.06

.22

-.05

-.41 .45 .23 .33

-.21 .27 .35

-.17 .42

-.51

.31

-.13

.23

.19



As shown in Figure 3.6, modelling analysis supports the initial model

proposed in Figure 3.2. In this model, Neuroticism predicts ASQ Negative

(standardized coefficient = .31). Conscientiousness predicts ASQ Positive with

coefficients of .22. As predicted, Internal Positive and Internal Negative are

positively correlated (r = .31), Stable Positive is positively correlated with Stable

Negative (r = .60). Similarly, Global Positive is positively correlated with Global

Negative (r = .64). All FFM dimensions are correlated with each other, with

Neuroticism negatively correlated with the four other FFM factors.

Multi-group SEM for testing gender differences of the model ASQ and FFM

Similarly, multi-group SEM was conducted to test gender differences of the model

for ASQ and FFM. See details in Chapter 3.3. This model was first tested in the male

group. Fit measures for this model indicated acceptable fit between model and data

(χ2 (22) = 29.71, p < 0.5; CFI = 0.98, TLI = 0.96, RMSEA = .052). Then, this model

was tested in the female group. For female, this model showed a good fit between

model and data (χ2 (22) = 29.18, p < 0.5; CFI = 0.99, TLI = 0.98, RMSEA = .032).

Finally, multi-group SEM was conducted to test gender differences of this

model. In addition to unconstrained base model, Measurement weights, Structural

covariances, and Measurement residuals were used as constrained conditions in multi

group analysis. The fit statistics for baseline comparisons of all models tested are laid

out in Table 3.12. It indicated that the unconstrained model fits best for the data.

Three constrained models have similar fits as the unconstrained model. Thus, this

model is compatible in both male and female in this sample.



Model NFI

Delta1

RFI

rho1

IFI

Delta2

TLI

rho2 CFI △CFI

Unconstrained .958 .895 .989 .971 .988


Structural residuals .906 .862 .957 .935 .956 -.032


Table 3.12: Baseline comparisons for tested models between ASQ and FFM.

3.4 Optimism and the Five-Factor Model of personality

The link between attributional style, dispositional optimism, and traditional

personality traits has great value in understanding both optimism constructs in a

broader area. Taking optimism as personality trait is also supported by its

considerable stability manifested in some genetic research mentioned earlier in

Chapter 1.

In the present study, examining correlations among dispositional optimism,

explanatory style, and the FFM factors provides some evidence of the related but

distinct relationship between these two optimism structures. Generally, both LOT-R

Optimism and ASQ Total have significantly negative correlations with Neuroticism,

and significantly positive correlations with Extraversion, which is consistent with

prior studies (e.g. Poropat, 2002; Sharpe et al., 2011). Specifically, both LOT-R

Optimism and attributional style for positive events had strong associations with four

of the five FFM factors, with the exception of Neuroticism, which is only

significantly correlated with LOT-R Optimism. On the other hand, three of the Big

Five factors, Neuroticism, Extraversion, and Conscientiousness showed strong

correlations with both LOT-R Pessimism and attributional style for negative events,

though Openness is only significantly correlated with LOT-R Pessimism and not



with ASQ Negative. The hypothesized correlation between LOT-R Pessimism and

Agreeableness was not significant. Similarly, the negative correlation between ASQ

Positive and Neuroticism was not found. As we predicted, ASQ Total was negatively

related to Neuroticism, and positively correlated with Extraversion, Agreeableness,

and Conscientiousness. The positive correlation between ASQ Total and Openness is

significant, though it has not been reported before.

In the comparison between correlations with specific facets of each Big Five

personality factor, dispositional optimism and explanatory style demonstrated mixed

patterns. For example, while LOT-R Optimism, LOT-R Pessimism, and ASQ

Negative are all strongly associated with depression, the correlation between

attributional style for positive events and depression didn’t reach statistical

significance. All these correlational patterns imply that explanatory style and

dispositional optimism are distinct but related constructs.

Do men and women have different patterns concerning the relationship between

optimism and FFM?

Gender differences in correlations between explanatory style and FFM have been the

focus of some prior studies. One such study conducted by Bunce and Peterson (1997)

reported that men and women were different in their attributional styles for negative

events and several personality traits, such as socialisation and good impression,

which were measured by the California Psychological Inventory (CPI). Also, Poropat

(2002) reported that the correlational patterns of attributional styles and FFM

dimensions appeared to have gender differences. However, correlational analyses

investigating potential gender differences in the relationship between dispositional

optimism and the Big Five personality factors have not been published previously as

to my knowledge. In my study, both the LOT-R and the ASQ scales were involved in

examining their associations with FFM dimensions for potential gender differences.

The correlational patterns observed in the current study were not quite

consistent with results in the study of Poropat (2002). Poropat reported that

Conscientiousness is correlated with ASQ Positive for women only and is correlated

with ASQ Negative for men only. However, results in the present study showed that



Agreeableness is the critical factor in differentiating men and women. Agreeableness

is correlated with ASQ Positive for men but not women, while it is correlated with

ASQ Negative for women but not men.

Considering the current study has been conducted in a Chinese sample while

Poropat (2002) collected data from a group of Austrian undergraduates, and no cross-

culture study regarding gender differences of the attributional style-FFM relationship

has been reported in prior literature, these different findings may due to cultural

influence. As regards potential gender influences on the relationship between LOT-R

scales and the main NEO-PI-R dimensions, results showed that Agreeableness is

correlated with LOT-R Pessimism for men but not for women. Openness is

correlated with both LOT-R Optimism and LOT-R Pessimism for women but not for

men.

Teasing apart dispositional optimism and dispositional pessimism by linking

them to the FFM

It has been proposed that dispositional optimism and dispositional pessimism have

distinct associations with the Big Five Personality factors, in which Neuroticism and

Extraversion play a larger role than the other three FFM factors (Marshall et al.,

1992). In Marshall et al.’s widely cited study, results indicated that dispositional

optimism correlated more strongly with Extraversion than did dispositional

pessimism, and dispositional pessimism showed a stronger correlation with

Neuroticism than did dispositional optimism, and thus also supported a two-factor

model of the LOT (Marshall et al., 1992).

Since then, this two-factor model has been demonstrated in many studies (Chang

et al., 1997; L. Chang & McBrideChang, 1996; Creed et al., 2002; Roysamb &

Strype, 2002). Based on previous research and the modelling analysis in Chapter 2.2,

an initial base model, which incorporates two differentiable factors (LOT-R

Optimism and LOT-R Pessimism) through their links to the FFM, was proposed. The

hypothesized model of the relationship between dispositional optimism and the FFM

is partially supported. Extraversion predicts both LOT-R Optimism and LOT-R

Pessimism, but Neuroticism influences only LOT-R Optimism. These results are in



agreement with previous findings that Extraversion and Neuroticism are the two

most influential predictors of optimism. The result that Neuroticism is not a predictor

of pessimism in this model is quite unusual considering the strong relationship

between these two variables in most previous studies.

Based on these findings, dispositional optimism may be best viewed as reflecting

two distinct traits, namely Dispositional Optimism and Dispositional Pessimism,

which are reflected in LOT-R Optimism items and LOT-R Pessimism items

respectively. Scoring and interpretation of the LOT-R should reflect this. Responses

should be scored separately for Dispositional Optimism and Dispositional Pessimism.

For most individuals, it is possible to identify them as being optimistic in an absolute

sense, because they agree with optimistic items (e.g. ‘I’m always optimistic about my

future’) and disagree with pessimistic items (e.g. ‘I rarely count on good things

happening to me’). Similarly, pessimists are people who agree with pessimistic items

and disagree with optimistic items.

Are attributions for positive and negative events predicted differently by the

FFM?

Though both explanatory style and the FFM have been taken as important

personality traits, very few studies have explored the relationship between these two

constructs and even fewer such studies have adopted the NEO-PI-R as a FFM

measure and used a SEM approach. In those rare studies, attributions for negative

events has been found to be negatively correlated with Conscientiousness

(Musgrave-Marquart et al., 1997). Correlational analyses between ASQ and FFM

dimensions support this finding. Moreover, we found that attributional style for

negative and positive events had different correlational patterns with the FFM. While

ASQ Negative is positively correlated with Neuroticism, and is negatively correlated

with Extraversion and Conscientiousness, ASQ Positive is positively related to four

of the five NEO-PI-R dimensions, excepting Neuroticism.

The hypothesized model of the relationship between attributional style and

the FFM was partially supported. As expected, Neuroticism predicts ASQ Negative.



A new relationship between Conscientiousness and ASQ Positive, which initially

wasn’t raised, emerged in this model. Previously, ASQ Negative has been reported to

be negatively correlated with Conscientiousness (Musgrave-Marquart et al., 1997;

Poropat, 2002). Though attributions for positive and negative events may reflect

differentiated cognitive styles, these results suggest that Conscientiousness may be

considered as one important FFM predictor of attributional style.

In the examination of the psychometric structure of the ASQ in Chapter 1,

results suggested that subjects apply consistent cognitive styles independent of event

valence, with personal tendencies to explain events as, for instance, global or local:

Subjects rating positive events as global tended also to describe negative events in

terms of global attributions, and likewise for the other two styles. These coherent

tendencies in cognitive styles are supported in the model, which links the ASQ and

FFM. Internal Positive and Internal Negative are positively correlated, so are the

other two cognitive style factors (Stability and Globality).


Chapter 4: Optimism and psychological well-being 117

Chapter 4: Optimism and psychological well-

being

4.1 Optimism and two approaches of well-being

Dispositional optimism and optimistic explanatory style have been taken as

theoretically connected. For instance, Scheier and Carver (1992) found that

differences in people’s expectations result in optimistic versus pessimistic

consequences. Also, Peterson and Seligman (1984) claimed that people’s attributions

for past events influence what they expect for the future. If individuals attribute past

failures to causes that are internal, lower self-esteem tends to be displayed and passive

expectation will be produced. If the explanation for a negative event is explained by

stable factors, individuals will expect more failures in the future, because the cause is

likely to remain for a long period. Similarly, if the cause of a negative event is

attributed to factors that are global, the expectations tend to be that these causes will

not be controllable even in different situations.

Empirical studies provide evidence for the link between explanatory style and

dispositional optimism. For example, one study revealed that individuals with positive

expectations for success also tend to have favourable attributions for their

performance (M. Marshall & Brown, 2006). Additionally, dispositional optimism and

explanatory style have also long been connected to each other because both variables

have been found to be closely correlated with depression, well-being, and other

related psychological constructs (Carver et al., 2010; Forgeard & Seligman, 2012;

Yuan & Zhang, 2007).

As optimism has been mainly conceptualized and measured in two constructs

of dispositional optimism and optimistic explanatory style, well-being has been



measured largely in two distinct traditions, of hedonic and eudemonic well-being.

While hedonic or subjective well-being relates mainly to happiness, the eudemonic

tradition focuses on psychological well-being, which is most widely implemented

using the Ryff scales of psychological well-being (RSPW; Ryff, 1989; Ryff & Keyes,

1995).

In the field of positive psychology, the study of psychological well-being,

which was developed by Ryff (1989), is very important, because this eudemonic

approach of well-being stems from personal development, the effort and desire to

achieve goals of life, and coping styles for life challenges. Six dimensions have been

identified in Ryff’s psychological well-being model, namely: self-acceptance or

positive attitudes toward oneself, personal growth or development, purpose in life,

control or mastery of the environment, positive relationships with others, and

autonomy or ability to be independent. These six dimensions present a set of

assessments related to positive performance, representing a general feeling of

happiness that are distinct from subjective well-being (Ryff & Singer, 2006). As one

of the most important predictors of well-being, optimism has been included in

numerous studies that examined well-being, though they mainly focused on subjective

well-being before the implementation of Ryff’s psychological well-being.

Dispositional optimism has been found to be positively related to

psychological well-being. For example, using an SEM approach, Augusto-Landa et al.

(2011) reported in a sample of 217 undergraduates that dispositional optimism

showed significant positive associations with all six psychological well-being

dimensions (r ranged from .38 to .59). Similarly, in a study conducted within a group

of 225 older adults, Ferguson and Goodwin (2010) found that dispositional optimism

was positively correlated with Purpose in Life (one of the six psychological well-

being dimensions; r = .46). The positive correlation between dispositional optimism



and psychological well-being has also been reported in an adolescent sample

(Monzani et al., 2014), with LOT-R scores positively correlated with all six

dimensions of the RSPW (r ranged from .32 to .56).

However, the relationship between explanatory style and psychological well-

being, which is measured by the RSPW, has not been reported to my knowledge.

Additionally, though there is much research suggesting that optimism is positively

associated with high levels of well-being (Scheier & Carver, 1992; Scheier. et al.,

2001), little has been done to explore the potential model of the two approaches of

optimism and psychological well-being in one single study. Because expectations are

regarded as a sufficient condition for maladaptive passivity following adversities

(Abramson et al., 1978), it is rational to infer that expectations may mediate the

relationship between explanatory style and well-being. Accordingly, it is reasonable

to construct a model in which explanatory style influences psychological well-being

through dispositional optimism.

Though explanatory style has not been linked to psychological well-being

previously, the mediating role of explanatory style between dispositional optimism

and subjective well-being has been examined in several previous studies. For example,

Isaacowitz (2005) reported that negative affiliated explanatory style and dispositional

optimism and pessimism predict subjective well-being (life satisfaction) measures

across three different age groups (280 young, middle-aged, and older adults). In one

study with a Chinese undergraduate sample (N = 350), Yuan and Zhang (2007)

reported that ASQ Total was negatively correlated with dispositional optimism (r = -

.30) and Satisfaction with Life (r = -.21) and positively correlated with depression (r

= .26). Dispositional optimism was revealed to be a mediating variable that mediates

the relationship between explanatory style and subjective well-being (depression and

Satisfaction with Life).



Currently, few investigations, however, have tested both dispositional

optimism and explanatory style together in the research of psychological well-being,

or examined the potential mediating role of expectations on the relationship between

attributional style and psychological well-being. There is no published research on the

relationship between attributions, expectations, and psychological well-being to my

knowledge.

In summary, previous investigations of optimism and well-being have shared

two primary limitations: first, they have exclusively assessed only one construct of

optimism (e.g. Augusto-Landa et al., 2011) or merely one approach of well-being (e.g.

Ahrens & Haaga, 1993). Second, even in studies where the two fundamental

constructs of optimism have both been assessed, research has not yet explored the

potential mediating model linking all these constructs. Therefore, my study aimed to

extend the positive psychology literature by examining the relationships among

dispositional optimism, explanatory style, and psychological well-being in a non-

Western sample. A further aim was to examine dispositional optimism as potential

mediator of the beneficial effects of optimistic explanatory style on psychological

well-being.

As an exploratory step, I first tested a model in which dispositional optimism

and dispositional pessimism were hypothesized to predict RSPW dimensions (see

Figure 4.1). In this model, LOT-R Optimism and LOT-R Pessimism are two

differentiated but negatively correlated factors. RSPW dimensions (correlated with

each other) will be predicted by LOT-R Optimism and LOT-R Pessimism. We next

tested a model constructing the predictive role of explanatory style on RSPW

dimensions (see Figure 4.2). In this model, ASQ Positive and ASQ Negative are

hypothesized to influence and predict RSPW dimensions (correlated with each other).




dispositional optimism and psychological well-being.


explanatory style and psychological well-being.

LOT-R Optimism

LOT-R Pessimism

LOT-R 10

LOT-R 4

LOT-R 1

Personal Growth

Positive Relations with Others

Purpose in Life

Self-Acceptance

Environmental Mastery

Autonomy

LOT-R 3

LOT-R 7

LOT-R 9

ASQ Positive

ASQ Negative

Global

Stable

Internal

Personal Growth


Purpose in Life

Self-Acceptance


Autonomy

Internal

Stable

Global



If the first two models are supported by the data, we will then examine a

model in which dispositional optimism acts as a potential mediator of the beneficial

effects of optimistic explanatory style on psychological well-being. This proposed

model, with LOT-R Optimism and LOT-R Pessimism partially mediating the effects

of ASQ Positive and ASQ Negative on psychological well-being, is shown in Figure

4.3.

Figure 4.3: Proposal for an initial model with hypothesized mediating role of

dispositional optimism between the relationship of explanatory style and


Since SEM analysis to examine the possible associations among explanatory

style, dispositional optimism, and psychological well-being has not been published

previously, alternative models of the relationships among these three variables will

also be explored. Specifically, the possibility that higher psychological well-being

may lead to more positive expectations as suggested by Ferguson and Goodwin (2010)

will be explored.

PWB

Autonomy


Personal Growth

Positive Relations with Others Purpose in Life

Self-Acceptance

Optimism

Pessimism

ASQ Positive

ASQ Negative



Correlational analyses will also be conducted. It is hypothesized that LOT-R

Optimism and ASQ Positive will be positively related to all RSPW dimensions, and

LOT-R Pessimism and ASQ Negative are expected to be negatively associated with

dimensions of psychological well-being.

4.2 Samples and instruments

Sample

Sample 1 was involved in the analysis of this study (for detail of this sample, see 1.5.4

of Chapter 1).

Instruments

Attributional style was assessed using the Chinese ASQ (Zhang, 2006). Composite

attributional styles were calculated separately for positive and negative events.

Reliabilities (Cronbach’s α) were acceptable 0.84 for the total and, for positive events

0.84; for negative events .77; for internality, .65; for stability, .76; and .80 for

globality.

Dispositional optimism was measured using a Chinese version of the Life

Orientation Test-Revised (Lai & Yue, 2000). Cronbach’sαfor the scale was 0.75; for

optimism, .79; and, for pessimism, .75.

Psychological well-being was measured with a Chinese version of the Ryff

Scales of Psychological Well-being (Chen, 2010). In the present sample, Cronbach’s

αcoefficients for the psychological well-being total was 0.92 (for self-acceptance, α

=.74; for positive relationships with other, α=.77; for personal growth, α=.78; for

purpose in life, α=.83; for environmental mastery, α=.81; for autonomy, α=.75).



Analysis strategy

Descriptive statistics and correlational analyses were first calculated. Structural

equation modelling (SEM) was then used to test a series of potential mediating

models constructing the relationships among explanatory style, dispositional

optimism, and psychological well-being using Amos 17.0 (Arbuckle, 2008). All

analyses took advantage of raw data supporting estimation of models using full

information maximum likelihood estimation.



(RMSEA). For CFI and TLI, values > 0.95 were taken as indicating acceptable fit (Hu

& Bentler, 1999). For the RMSEA, values of < .05 indicated acceptable fit (C. Y. Yu,

2002). Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC)

are reported to aid model comparison.

Criterion for mediating model

Four conditions must be met to establish an acceptable mediating model (Baron &

Kenny, 1986). First, the predictor variable (explanatory style) is related to the

outcome variable (psychological well-being). Second, the predictor variable

(explanatory style) is related to the potential mediator (dispositional optimism). Third,

the mediating factor (dispositional optimism) is related to the outcome variable

(psychological well-being). Finally, the relation between the predictor variable

(explanatory style) and the outcome variable (psychological well-being) significantly

decreases once the mediator (dispositional optimism) is included in the model.



4.3 Results


Table 3.1 shows the means, standard deviations, and Cronbach’s alpha of the total

samples. Reliabilities were acceptable.


LOT-R Optimism 8.27 1.84 0.79

LOT-R Pessimism 3.85 1.99 0.75

LOT-R Total 16.42 3.01 0.75

ASQ Negative 12.90 1.78 0.84

ASQ Positive 15.28 1.91 0.77

ASQ Total 2.38 2.17 0.84

RSPWS1 33.05 5.46 0.75

RSPWS2 37.65 5.60 0.81

RSPWS3 41.77 5.31 0.78

RSPWS4 40.06 6.45 0.77

RSPWS5 38.45 6.36 0.83

RSPWS6 34.70 5.80 0.74

RSPW Total 225.68 25.82 0.92

Table 3.1: Means, standard deviations and Cronbach’s alpha for all measures.

Note: Means for LOT-R dimensions are on a scale ranging from 1 to 5, ASQ

dimensions range from 1 to 7, and RSPW dimensions from 1 to 6, with higher

numbers indicating greater amounts of these qualities. RSPWS1, autonomy; RSPWS2,

environmental mastery; RSPWS3, personal growth; RSPWS4, personal relations with

others; RSPWS5, purpose in life; RSPWS6, self-acceptance (n = 452).



Correlational analyses

The first hypothesis tested was that explanatory style, dispositional optimism, and

psychological well-being would correlate positively and significantly with each other.

Table 3.2 shows the inter-correlations among the variables of interest.

As shown in Table 3.2, dispositional optimism was positively correlated with

explanatory style for positive events and all RSPW dimensions; dispositional

pessimism was negatively correlated with explanatory style for positive events and all

RSPW dimensions; dispositional optimism was negatively correlated with

explanatory style for negative events; and dispositional pessimism was positively

correlated with explanatory style for negative events. Explanatory style for positive

events was positively associated with all RSPW dimensions, while explanatory style

for negative events was negatively associated with all RSPW dimensions. Finally, all

RSPW dimensions were positively and significantly related to each other.



Measures LOT-R

Optimism

LOT-R ASQ

Negative

ASQ

Positive

ASQ RSPW

S1

RSPW

S2

RSPW

S3

RSPW

S4

RSPW

S5

RSPW

S6 Pessimism Total

LOT-R

Optimism -

LOT-R

Pessimism -0.24 ** -

ASQ Negative -0.13 ** 0.11 * -

ASQ Positive 0.15 * -0.18 ** 0.31 ** -

ASQ Total 0.23 ** -0.25 ** -0.54 ** 0.63 ** -

RSPW S1 0.28 ** -0.12 ** -0.27 ** 0.03 0.25 ** -

RSPW S2 0.37 ** -0.31 ** -0.33 ** 0.13 ** 0.39 ** 0.44 ** -

RSPW S3 0.28 ** -0.32 ** -0.12 ** 0.18 ** 0.26 ** 0.30 ** 0.44 ** -

RSPW S4 0.37 ** -0.35 ** -0.23 ** 0.11 * 0.29 ** 0.30 ** 0.55 ** 0.54 ** -

RSPW S5 0.28 ** -0.37 ** -0.21 ** 0.13 ** 0.29 ** 0.38 ** 0.47 ** 0.55 ** 0.50 ** -

RSPW S6 0.46 ** -0.28 ** -0.25 ** 0.15 ** 0.33 ** 0.41 ** 0.58 ** 0.33 ** 0.52 ** 0.44 ** -

RSPW Total 0.46 ** -0.40 ** -0.32 ** 0.16 ** 0.41 ** 0.63 ** 0.78 ** 0.71 ** 0.78 ** 0.76 ** 0.74 **

Table 3.2: Correlations between measures.

Note: RSPWS1, autonomy; RSPWS2, environmental mastery; RSPWS3, personal growth; RSPWS4, personal relations with others; RSPWS5,

purpose in life; RSPWS6, self-acceptance. (n = 452)

* p < 0.05. ** p < 0.01.



Structural Equation Modelling

First, the proposed model in which dispositional optimism predicts RSPW factors (as

shown in Figure 4.1) was tested.

Figure 4.4: Standardized estimations for the initial model of dispositional optimism

and psychological well-being.

For the initial base model, the fit was adequate (χ² (33) = 95.48 p < .001; CFI

= 0.961; TLI = 0.919; RMSEA = 0.063; AIC = 181.406; BIC = 370.635).

Standardized estimates of the original model are shown in Figure 4.4. Modifications

were suggested that significantly improved model fit, and the resultant model fit

reasonably well by all criteria (χ² (28) = 44.80, p = .023; CFI = 0.988; TLI = 0.973;

RMSEA = 0.036; AIC = 144.802; BIC = 350.486), as shown in Figure 4.5. The new

LOT-R Optimism

LOT-R Pessimism

LOT-R 10

.47 LOT-R 4

.71

LOT-R 1 .20

LOT-R 9 .76

LOT-R 7 .67

LOT-R 3

.28

Personal Growth


Purpose in Life

Self-Acceptance


Autonomy

.51

.50

.27

.43 .29 .69

-.08

-.33

-.27 -.28

-.16 .05

.25

.27

.39

.30

.22

.17

.08

.27

.10

.34

.28

.31 .42

.09

.23

-.32



paths all had loadings of .24 or below, suggesting deviation from the theoretical

model is minor (see Figure 2.4).

As shown in the figure, most direct paths in this model are significant except

the path between LOT-R Pessimism and Autonomy, between LOT-R Pessimism and

Environmental Mastery, and between LOT-R Pessimism and Self-Acceptance. As

predicted, LOT-R Optimism and LOT-R Pessimism are negatively correlated (r = -

.29). Six RSPW dimensions are predicted by LOT-R Optimism and three RSPW

dimensions are predicted by LOT-R Pessimism.

Figure 4.5: Standardized estimations for the modified model of dispositional

optimism and psychological well-being.

LOT-R Optimism

LOT-R Pessimism

LOT-R 10 .48

LOT-R 4 .70

LOT-R 1 .19

LOT-R 9 .75

LOT-R 7 .67

LOT-R 3

.30

Personal Growth


Purpose in Life

Self-Acceptance


Autonomy

d4

.56

.52

.23 .45

.31 .70

-.11 -.33

-.28 -.29

-.18 .03

.21

.29

.41

.29

.19

.18

.03

.24

.02

.32

.27

.28 .43

.12

.20

-.29 -.24

-.24

-.16

.16



I next tested the proposed model in which explanatory style predicts

psychological well-being factors (as shown in Figure 4.2). For the initial base model,

the fit was adequate (χ² (30) = 82.44 p < .001; CFI = 0.970; TLI = 0.934; RMSEA =

0.062; AIC = 178.439; BIC = 375.896). Standardized estimates of the original model

are shown in Figure 4.6.

Figure 4.6: Standardized estimations for the initial model of explanatory style and


Although CFI or GFI values may be considered adequate, modifications were

suggested and made to the original model to obtain a better fit according to the results.

These modifications include five relationships between the residual variances of

measured variables, for instance a relationship between the residual variance of

Autonomy and ASQ Positive Global. The new paths all had loadings of .23 or below,

suggesting deviation from the theoretical model is minor (see Figure 2.4). These

modifications significantly improved model fit, and the resultant model fit reasonably

well (χ² (25) = 41.31, p =.021; CFI = 0.991; TLI = 0.975; RMSEA = 0.038; AIC =

147.310; BIC = 365.336).

ASQ Positive

ASQ Negative

Global .57

Stable .83

Internal .69

.67

.55 .35

Personal Growth


Purpose in Life

Self-Acceptance


Autonomy .09

.21

.25 .18

.19 .22

-.36

-.28 -.31 -.22

-.47 -.34

.34

.36

.49

.43

.35

.24

.21

.31

.32

.47

.38 .48

.50

.24 .45

.28

.42

.61 Internal

Stable

Global



In this model, as predicted, three cognitive style factors (internality, stability,

and globality) are correlated with event valences. ASQ Positive and ASQ Negative

predict RSPW dimensions (except ASQ Positive and Autonomy).

Figure 4.7: Standardized estimations for the modified model of explanatory style and


Finally, I tested the preferred model, in which dispositional optimism acts as a

potential mediator of the beneficial effects of optimistic explanatory style on

psychological well-being (as shown in Figure 4.3). For the initial base model, CFI or

GFI values may be considered acceptable (χ² (29) = 132.558, p < .001; CFI = 0.920;

TLI = 0.875; RMSEA = 0.089; AIC = 184.558; BIC = 291.514). Standardized

ASQ Positive

ASQ Negative

Global .57

Stable .81

Internal .72

.77

.53

.31

Personal Growth


Purpose in Life

Self-Acceptance


Autonomy

.09

.22

.25 .20

.21 .23

-.34 -.26 -.29

-.23 -.43

-.34

.34

.35

.49

.44

.36

.23

.20

.31

.32

.47

.38

.49 .50

.23

.45

.39

.48

.45

-.05

.09

.12

.23 .23

Internal

Stable

Global



estimates of the original model are shown in Figure 4.8. Modifications were suggested,

which significantly improved model fit, and the resultant model fit reasonably well by

all criteria (χ² (23) = 37.88, p = .026; CFI = 0.988; TLI = 0.977; RMSEA = 0.038;

AIC = 101.880; BIC = 233.518) (see Figure 2.4).

Figure 4.8: Standardized estimations for the initial meditating model.

As shown in Figure 4.9, all direct and indirect paths in this model are

significant. This final modified model has a highly significant indirect path from

explanatory style to dispositional optimism to psychological well-being. Additionally,

the relationship between the predictor variable (ASQ Positive and ASQ Negative) and

the outcome variable (psychological well-being) (r = .32 and r = -.47, respectively)

significantly decreases (r = .17 and r = -.35, respectively) once the mediator (LOT-R

Optimism and LOT-R Pessimism) is included in the model. Thus the relationship

between explanatory style and psychological well-being was partially mediated by

dispositional optimism as originally proposed.

PWB

Autonomy

Environmental Mastery Personal Growth


Purpose in Life

Self-Acceptance

Optimism

Pessimism

ASQ Positive

ASQ Negative

.21

.18

.18

-.33 -.19

-.24

-.28

.37

-.19

.52

.76 .62

.74 .67

.71

.31



Figure 4.9: Standardized estimations for the modified meditating model.

4.4 Positive relationship between optimism and

psychological well-being

My study provided empirical evidence of the correlational patterns between

explanatory style, dispositional optimism, and psychological well-being in a non-

Western sample. Both dispositional optimism and explanatory style are strong

predictors of psychological well-being. The relationship between explanatory style

and psychological well-being, however, is predominantly mediated by dispositional

optimism and dispositional pessimism. The results were consistent with findings of

previous research in Western samples. That is, explanatory style and dispositional

optimism were weakly correlated (Forgeard & Seligman, 2012), but both of these two

PWB

Autonomy


Personal Growth


Purpose in Life

Self-Acceptance

Optimism

Pessimism

ASQ Positive

ASQ Negative

.20

.19

.17

-.35 -.19

-.23

-.30

.33

-.20

.61

.77 .58

.73 .65

.67

.17

.21

.31

-.18

-.13

.20

.27



constructs of optimism were moderately correlated with well-being (Carver et al.,

2010).

Positive relationships were found between LOT-R Optimism and

psychological well-being dimensions. More optimistic individuals reported a higher

level of PWB, which is consistent with studies conducted in Western participants.

That is, individuals who have positive expectation for the future are more likely to

report high levels of psychological well-being. There is evidence that optimists can

cope more adaptively with stress and, therefore, gain psychological benefits (Scheier

& Carver, 1992). Similar results have been found in other studies (Carver et al., 2010).

Inversely, negative correlations were found between LOT-R Pessimism and

dimensions of psychological well-being. These findings correspond with results

reported by Chang et al. (1997) and Mäkikangas and Kinnunen (2003).

Consistent with previous studies that individuals who have an optimistic

explanatory style are more likely to report higher levels of psychological well-being

than people with a pessimistic attributional style (Wise & Rosqvist, 2006), the current

results revealed that scores on attributions for positive events were positively

correlated with levels of all six dimensions of psychological well-being. Optimists are

believed to face adversity and deal with negative situations more effectively than

pessimists and, therefore, gain more psychological benefits. Optimistic explanatory

style may serve as a protective factor for well-being. Additionally, dispositional

optimism was positively correlated with explanatory style, which is consistent with

some previous studies exploring the relationship between these two constructs.

The most important goal of the current study was to address whether

dispositional optimism mediated the link between explanatory style and psychological

well-being. The proposed mediating model was tested and supported. It indicated that

an optimistic explanatory style was a strong predictor of psychological well-being, as



measured by the RSPW. However, the effect of explanatory style on psychological

well-being was mediated by dispositional optimism as shown in the mediating model.

Thus, this study provides conditional evidence for the mediating role of dispositional

optimism in the relationship between attributional style and psychological well-being.

Myers and Diener (1995) suggested that the causal direction from traits to

subjective well-being may be reversed. It might be similar for psychological well-

being. Given the cross-sectional nature of these findings, the causal directions

depicted in these models may be the reverse of what was predicted. Higher levels of

psychological well-being, such as positive relations with others, may contribute to

positive expectations. However, no empirical evidence with longitudinal studies for

these reversed patterns has been carried out, as far as we know. Thus, despite a good

statistical model fit for some models with pathways from psychological well-being to

optimism (tested but not reported in Results), these models are less plausible than the

final resultant meditating model, due to lack of evidence.

Overall, this study provided consistent evidence of, and further support for, the

beneficial effects of both types of optimism on psychological well-being in a college

student sample. Both dispositional optimism and optimistic explanatory style are

strong predictors of psychological well-being. While both dispositional optimism and

explanatory style have a direct effect on psychological well-being, the effect of

explanatory style on psychological well-being was partially mediated by dispositional

optimism in the final model. It is valuable to note that an optimistic explanatory style

clearly contributes to enhancing individuals’ psychological well-being.


Chapter 5: Cultural influence on optimism 136

Chapter 5: Cultural influence on optimism

5.1 Cultural issues: from the West to the East

Research shows that optimism as a whole has adaptive value in dealing with

environmental risks and life challenges over the million or so years of evolution

(Tiger, 1979). This adaptive advantage still works for people to achieve more in

current life (Carver et al., 2010; Seligman, 2011). The universality of being

optimistic (Michalos, 1988) and the prevalent positive associations among optimism,

subjective well-being, and perceived physical health (Gallagher, Lopez, & Pressman,

2013), have been known for a long time.

Though benefits of being optimismtic are widely acknowledged, a crucial but

often neglected concern in studying optimism is the examination of this important

psychological concept across different cultural and ethnic groups. Optimism-related

studies in recent years have been mainly conducted in Western cultures particularly,

so the results do not necessarily apply to behaviours in other cultures. Is there any

cultural difference concerning optimism-related properties? The answer may not be

as simple as it seems. To make it clear scientifically, empirical studies must be

carried out to examine whether cultural differences have considerable and

meaningful effects on optimism. The following study set out to address this question

and examine group differences on measures of dispositional optimism and

explanatory style between Eastern and Western cultures.

It is assumed that most Eastern societies, such as those in China and India,

miantain a collectivist or an interdependent self, whereas most Western societies,

such as the U.S. and Canada, foster an individualistic or an independent self (Markus

& Kitayama, 1991). These conceptions of self, in turn, may relate to an individual’s

explanation for events in their life and generalised different expectations for their

future. To be specific, one of the distinctions between Eastern and Western cultures

concerns the level of separation between the achivement domain and the

interpersonal domain in life events (Higgins & Bhatt, 2001). It has been assumed that

individuals from a collectivist culture may not differentiate these two domains as

sharply, due to a lack of separation of self from others (Higgins & Bhatt, 2001).



To understand the influence of culture on optimism, it is critical to review

recent findings associated with the examination of optimism between these two

cultures.

5.2 Prior studies investigating cultural differences in optimism

Optimism studies conducted in both cultures

Within the broad and divergent culture frames of the East and West, differences in

both dispositonal optimism and attributional styles have been examined by

researchers from an cross-cultural perspective. J. G. Miller (1984) carried out one of

the earlier studies about cultural influences on explanatory style within a group of

Hindus and a group of Americans. He (or she) found that individuals in Western

cultures emphasised the role of internal factors in causal explanations of events,

whereas individuals in Eastern cultures tended to view the external factors as playing

a determining role in causing various life events.

Lee and Seligman (1997) also investigated cultural influences on causal

attributions. A sample of 257 white American undergraduates, a group of 312

mianland Chinese college students, and 44 Chinese-American students (32 subjects

were American-born Chinese, the others were non-American-born Chinese but had

stayed in the United States for 5.5 years on average) were recruited and completed

the ASQ. The authors found that the White Americans had a more optimistic

explantory style than Chinese-Americans, and Chinese-Americans were

characterized with a more positive attributional style than mainland Chinese. Using

the same scale, Higgins and Bhatt (2001) conducted a cross-cultural study within

Indian (n = 195) and Canadian (n = 162) college students. They found that Indian

students generated more contextual attributions for life events than did the Canadian

students.

As discussed earlier in Chapter 1, attributional style has been examined along

still another line – attributional bias, which overlaps with both the definition and



measurement of optimistic explanatory style. According to Higgins and Bhatt (2001,

p. 55), both Westerners and Easterners showed “a self-serving tendency to explain

negative events with external-uncontrollable causes and to explain positive events

with internal-controllable causes”. That is, both cultures showed an attributional bias,

generating more external, uncontrollable causes to explain negative events and more

internal, controllable causes to explain positive events.

This self-serving attributional bias, or an optimistic explanatory style, has

been previously studied with a cross-culture perspective, and cultural effects were

reported (e.g., Kashima & Triandis, 1986). For example, the study of Lee and

Seligman (1997) indicated that Mainland Chinese attributed their success to others or

circumstances and their failure to themselves more often than did White Americans.

This idea was supported in a meta-analysis of 266 studies, including subjects from

different cultural background. Mezulis et al. (2004) reported that Asian samples

generally displayed significantly smaller attributional bias than U.S. or Western

samples. That is, Westerners received higher scores on optimistic explanatory style

than Easterners.

These studies came to the conclusion that individuals from Eastern cultures,

or so-called collectivistic cultures, expressed less self-serving attributional bias than

individuals from the West, or individualistic cultures (e.g., Higgins & Bhatt, 2001).

This finding was consistent with traditional cultural differences that Westerners have

more self-serving bias than Easterners (Lee & Seligman, 1997). However, there are

discrepancies in the level of self-serving attributional bias even among countries with

similar cultural backgrounds. For example, while both Americans and Finnish people

showed a tendency to apply self-serving bias in attribution, American participants

expressed a greater bias than Finnish subjects (Nurmi, 1992).

Using the dispositional optimism framework, Chang and colleagues

investigated the potential mechanism underlying cultural influences on optimism and

pessimism for Westerners and Easterners (Chang, 1996; Chang, Sanna, & Yang,

2003). In one of their earlier studies (Chang, 1996), 111 Asian-American and an

equal number of White American students completed an adapted version of the



original LOT. The authors found that Asian-Americans scored significantly higher

on pessimism than White-Americans, which was consistent with traditional images

of Western-Eastern cultural differences. The results were partly replicated in another

cross-cultural study by Sinha, Willson, and Watson (2000). College students from

India (n = 198) and Canada (n = 344) were assessed on their level of dispositional

optimism and several other psychological factors. The authors found that Indian

students were more pessimistic than their Canadian counterparts.

Abdel-Khalek and Lester (2006) compared levels of dispositional optimism

of Kuwaiti (n = 460) and American (n = 273) college students using an adapted

version of the original LOT. Consistent with findings of Chang et al. (2003), the

Easterners scored significantly higher on pessimism than their Western counterparts.

However, they also found that Kuwaiti students were less optimistic than American

students, which was not found in the study of Chang et al. (2003).

Cultural differences in optimism have been supported by some meta-analytic

studies as well. For example, Nes and Segerstrom (2006) investigated the potential

differences in optimism and coping between English-speaking and non-English-

speaking countries. Looking at 50 studies, they found that participants involved in

studies in the United States or in English-speaking countries showed stronger

correlations between dispositional optimism and coping strategies than did

participants from non-English-speaking nations.

Other studies have investigated age-related dipositional optimism across

different cultures. For example, in samples including Americans and Hong Kong

Mainland Chinese, You, Fung, and Isaacowitz (2009) reported that older Mainland

Chinese displayed a lower level of dispositional optimism than did younger

Mainland Chinese, whereas older Americans showed a higher level of dispositional

optimism than their younger counterparts.

To summarize the findings to date, from the perspective of dispositional

optimism, it is generally agreed that Westerners are more optimistic than Easterners.

However, there is at least one exception. Chang et al. (2003) investigated the cultural



influence on the role of optimism in predicting life satisfaction and depressive

symptoms. A sample of 294 South Korean and 320 European-American

undergraduates were tested on optimism, depression and subjective well-being.

Surprisingly, the South Korean students were found to be significantly less

pessimistic than the European-American students. No significant group difference on

levels of optimism between these two ethic groups was found.

The author stated that his findings were consistent with his earlier studies

conducted between Asian Americans and European Americans. However, these

groups were not strictly comparable. Further research is necessary to continue to

explore the possibility of discrepancy between specific ethnic groups. Though

research based on explanatory style has generally found that Westerners are more

optimistic than Easterners, cultural comparisons in attributional style have led to

mixed results, which suggest that cultural influences on explanatory style is not

always consistent, at least for some dimensions.

Optimism studies conducted in Easterners

In addition to cross-cultural studies that directly compare the differences in optimism

expression between Eastern and Western cultures, optimism-related research recently

conducted only within Eastern cultures has provided some findings for better

understanding of both dispositional optimism and explanatory style.

Yu and Seligman (2002) investigated associations between explanatory style

and levels of depressive symptoms and other variables within a group of Chinese

children (n = 185). The study replicated previous findings that pessimistic

explanatory style was negatively associated with academic achievement and

positively correlated with school conduct problems. Additionally, in their optimism

intervention study conducted in a Chinese sampe of 220 students with depressive

symptoms, the intervention group showed significantly fewer depressive symptoms

than the control group, and this benefit continued at 3- and 6- month follow-ups.

More studies have been conducted in dispositional optimism than in

explanatory style in Eastern cultures. One study using a Taiwanese sample (n = 381)



examined the potential mediating role of social support between dispositional

optimism, subjective well-being (happiness and life satisfaction), and psychological

well-being (personal growth and purpose in life). The authors found that dispostional

optimism was positively associated with both subjective well-being and

psychological well-being, which had been supported by many previous studies

conducted in Western cultures (Tseng, 2007).

In another study, Ho, Cheung, and Cheung (2010) examined the role of

optimism in promoting subjective well-being within 1,807 adolescents in Hong Kong.

It showed that dispositional optimism was positively associated with life satisfaction

(r = .48, p < .05) and was negatively associated with psychosocial problems (r = -.72,

p < .05), which were consistent with previous findings in Western cultures (Wrosch

& Scheier, 2003). Also, with a sample of 250 community-dwelling older Koreans, Ju,

Shin, Kim, Hyun, and Park (2013) assessed the level of dispositional optimism,

Meaning in Life and subjective well-being of the participants. The authors found that

dispositional optimism was positively associated with both subjective well-being (r

= .50) and meaning in life (r = .75) in one group of old adults.

5.3 The present study

Previous studies revealed cultural influences on different optimism expressions

between Eastern and Western cultures, though some results were inconsistent.

Because most published research in cultural differences on optimism has been

conducted between Americans and some Eastern nations, and there are no published

studies that have compared cultural influences on optimism between British White

people and Eastern countries, we know very little about the potential cultural

influence on optimisim within these two ethnic groups. Therefore, the goal of the

present study was to extend the optimism literature by examining the differences in

dispositional optimism and explanatory style between Mainland Chinese and British

White people.



The main purposes of the present study were to (1) test whether the ASQ and

the LOT scored within a group of White British people possess the same

psychometric structures as explicated in the sample of Mainland Chinese in Chapter

2; (2) examine correlations between measures of dispositional optimism and

explanatory styles among Easterners (Mainland Chinese) and Westerners (British

White); (3) assess potential group differences on measures of dispositional optimism

and explanatory styles between the two ethnic groups.

In agreement with the long-held perspective on cultural differences between

Easterners and Westerners, it was expected that measures of dispositional optimism

and explanatory style would be significantly intercorrelated with each other for both

cultural groups. In addition, it was expected that both Mainland Chinese and British

White groups would show an optimistically-biased attributional style, generating

more external, unstable, and specific causes to explain negative events and more

internal, stable and global causes to explain positive events. However, the

relationship between these variables may not be identical given cultural differences

between Easterners and Westerners. I did not generalize specific hypotheses

regarding levels of pessimism and explanations since results in prior research were

inconsistent, and the current study is the first to examine potential cultural

differences on optimism between these two groups.

Modelling Analyses and analysis techniques

We first tested the ASQ model (three-factor model of negative events and positive

events) described in Chapter 2 in the White British sample; and then replicated the

two-factor model of the LOT-R described in Chapter 2 in the Western participants.

Structural equation modelling (SEM) was used to test these models using Amos

17.0 (Arbuckle, 2008). All analyses took advantage of raw data supporting

estimation of models using full information maximum likelihood estimation. The

adequacy of model fit was assessed using the comparative fit index (CFI), Tucker-

Lewis index (TLI) and the Root Mean Square Error of Approximation (RMSEA).

For CFI and TLI, values > 0.95 were taken as indicating acceptable fit (Hu & Bentler,



1999). For the RMSEA, values of < .05 indicated acceptable fit (C. Y. Yu, 2002).

Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) are

reported to aid model comparison.

5.3.1 Method

Participants

Data were collected from undergraduates in Mainland China and the United

Kingdom. The Mainland Chinese sample consisted of 232 undergraduates in Sample

2. A total of 205 White British participants were included in Sample 3. See 1.5.4 of

Chapter 1 for details of these two samples.

Materials

The original English version of the Life Orientation Test-Revised (LOT - R; Scheier

et al., 1994) was used to measure dispositional optimism in the UK sample. A

Mainland Chinese version of Life Orientation Test-Revised (CLOT-R; Lai et al.,

1998) was used to measure dispositional optimism of the Mainland Chinese students.

The original English version of the ASQ (Peterson et al., 1982) was used to

measure explanatory style of the UK students. Attributional Style of Mainland

Chinese participants was measured using a Mainland Chinese version of the ASQ

(Zhang, 2006).

Procedure

For the Mainland Chinese sample, participants were tested in groups of 30 to 50 by

their teacher. Each teacher was trained on the administration of the task. After

detailed instructions were provided, participants completed the paper-and-pencil

questionnaires. Testing took around 20 minutes.

For the White British sample, two measures were administered to all 205

participants as part of one bigger survey that was completed in the form of paper-

and-pencil questionnaires. Instructions to all participant groups were identical. Of the

initial White British sample, three participants provided an incomplete set of surveys,



and thus left a total of 202 completed responses that were available for subsequent

data analyses.

5.3.2 Results


We first examined descriptive and summary statistics, and the standard composite

explanatory style scores. Table 5.1 shows the descriptive statistics of the ASQ and

the LOT-R in both groups. Reliabilities were acceptable.

Measures

Culture group

Mainland Chinese White British

Means SD Cronbach’s

Alpha Means SD

Cronbach’s

Alpha

ASQ Total 2.12 2.24 0.83 1.26 2.21 0.83

ASQ Negative 12.98 1.92 0.79 12.21 1.83 0.78

ASQ Internal Negative 4.47 0.61 0.40 4.34 0.78 0.61

ASQ Stable Negative 4.33 0.89 0.69 4.03 0.83 0.70

ASQ Global Negative 4.18 0.96 0.73 3.84 0.89 0.74

Hopelessness 4.25 0.81 0.80 3.94 0.76 0.81

ASQ Positive 15.1 1.81 0.81 13.47 1.94 0.81

ASQ Internal Positive 4.87 0.69 0.63 4.58 0.88 0.71

ASQ Stable Positive 5.29 0.78 0.71 4.63 0.82 0.71

ASQ Global Positive 4.94 0.8 0.63 4.26 0.79 0.60

Hopefulness 5.12 0.69 0.78 4.45 0.7 0.76

LOT-R Optimism 8.37 1.93 0.46 7.02 2.38 0.57

LOT-R Pessimism 4.05 2.23 0.64 4.51 2.15 0.68

Table 5.1: Means, SDs and Cronbach’s Alpha for the ASQ and the LOT-R scales.

Note: For Mainland Chinese, N=232. For White British, N=202. Correlations inside

of parentheses are for White British. Hopelessness = stability + globality of the ASQ

negative events; Hopefulness = stability + globality of the ASQ positive events.



Modelling

Following the study of Hewitt et al. (2004) and our analysis in Chapter 1, method

(event) variance was accommodated using an MTMM structure in all modelling

analysis. We first tested the hypothesis that the structure of explanations for the

causes of negative events reflects three factors of internality, stability and globality

which are correlated based on data of Sample 3.

The base model without modifications did not fit very well (χ² (114) = 217.19,

p < .001; CFI = 0.88; TLI = 0.82; RMSEA = 0.067; AIC = 331.19; BIC = 519.77).

After modifications, the fit was improved by all criteria (χ² (98) = 122.38, p <.05;

CFI = 0.97; TLI = 0.95; RMSEA = 0.035; AIC = 268.38; BIC = 509.88). In this

modified model, internality and stability factors correlated .20; stability and globality

had an r of .66, internality and globality was uncorrelated (r = -.01). Thus, the data

collected from Sample 2 didn’t support the model previously reported by Hewitt et al.

(2004) and the similar model found in Chapter 1. Here the corrected model of causal

attributions for negative events emerged as different correlations between three

factors (correlated internality-stability and correlated globality-stability but

uncorrelated internality-globality). We next turned to see if this model would fit well

for positive events.

A model for positive events was constructed in the same fashion as the

baseline model for negative events. Fit measures for this model indicated a lack of

adequate fit between model and data (χ2

(114) = 198.15, p < 0.001; CFI = 0.91, TLI =

0.88, RMSEA = 0.061; AIC = 312.15; BIC = 500.72). But modifications were

suggested and these modifications improved fit by all criteria (χ² (104) = 133.66, p

<.05; CFI = 0.97; TLI = 0.96; RMSEA = 0.038; AIC = 267.66; BIC = 489.31). In the

correlated factor model stability and globality correlated .63, internality and

globality .34 and internality and stability .59 (See Figure 5.1).



Figure 5.1: Well-fitting 3-factor model of attributional style for positive events.

As a result, as previously reported by Higgins et al. (1999) and in Chapter 1,

a model of causal attributions for positive events in terms of three correlated factors

of globality, stability, and internality adequately accounted for responses to these

positive events in the ASQ.

Analyses of separate ASQ positive events and ASQ negative events, then,

indicated that only ASQ scale of positive events was well accounted for by three

INTERNALITY

STABILITY

GLOBALITY

A1.12.1 d6

.71 A1.10.1 d5

.69 A1.9.1 d4

.48 A1.6.1 d3 .72

A1.3.1 d2 .46

A1.1.1 d1

.28

A1.12.2 d12

.72 A1.10.2 d11 .75

A1.9.2 d10 .57

A1.6.2 d9 .67

A1.3.2 d8 .43

A1.1.2 d7 .25

A1.12.3 d18

.65 A1.10.3 d17

.59 A1.9.3 d16

.64 A1.6.3 d15

.51

A1.3.3 d14 .31

A1.1.3 d13 .04

.59

.63

.34

.23

.09

.40

.19

.33

.29

.29

.39

.08

.14

.24

.27

.25

.25

.21

.24

.21

.37

.21

.17 -.15

.19

-.19

-.18

-.23 -.20

.14

.15



correlated factors of internality, stability, and globality. ASQ scale of negative events

didn’t support this three correlated-factor model.

Testing for measurement invariance of ASQ-Positive across cultures

Modelling analysis shows that only ASQ scale of positive events was well accounted

for by three correlated factors of internality, stability, and globality across two

cultures and ASQ scale of negative events didn’t support this three correlated-factor

model in the White British sample. Thus, to test measurement invariance of ASQ,

only ASQ scale of positive events was tested using multi-group SEM. In addition to

unconstrained base model, Measurement weights, Structural covariances, and

Measurement residuals were used as constrained conditions in multi group analysis.

The fit statistics for baseline comparisons of all models tested are laid out in Table

5.2. Table 5.2 shows that the unconstrained model fits best for the data. Three

constrained models have similar fits as the unconstrained model. Thus, ASQ-Positive

model is identical in measuring attributional style across two cultures.

Model NFI

Delta1

RFI

rho1

IFI

Delta2

TLI

rho2 CFI △CFI

Unconstrained .842 .788 .934 .908 .931

Measurement weights .836 .793 .934 .914 .932 . 001



Table 5.2: Baseline comparisons for tested ASQ-Positive models

Structural equation modelling for the LOT-R

We first test the one-factor model; all six items were specified as indicators of a

single factor. The unidimensional model fit poorly with the data, with (χ² (10) =



123.83, p < .001; CFI = 0.343; TFI = 0.015; RMSEA = 0.238; AIC = 145.828; BIC =

146.622).

Figure 5.2 Standardized estimations for the two-factor model

We next turn to the two-factor model. Here the three positively worded items

were specified as indicators of the Dispositional Optimism factor, and the three

negatively worded items were specified as indicators of the Dispositional Pessimism

factor. Compare with the one-factor model, the two-factor model fit much better with

χ² (8, N = 202) = 21.387, p < .005; CFI = 0.923; TFI = 0.855; RMSEA = 0.091; AIC

= 47.387; BIC = 90.394). From the modified index, we established relationships

between the residual variance of Item 1 and Item 7, and between the residual

variance of Item 1 and Item 9. These modifications improved fit by all criteria (χ² (6)

= 6.86, p <0.5; CFI = 0.995; TLI = 0.988; RMSEA = 0.027; AIC = 36.860; BIC =

86.484). The correlation between the Dispositional Optimism factor and the

Dispositional Pessimism factor was -.27 (p<.01). The factor loading ranged from .30

to .81 (See Figure 5.2).

Dispositional Optimism

.77

LOT-R 10

LOT-R 4 .56

LOT-R 1 .38

Dispositional Pessimism

.75

LOT-R 9

LOT-R 7 .68

LOT-R 3 .52

-.27 .12

-.25



Thus, as previously reported by many studies conducted in Western cultures,

a two-factor model of dispositional optimism was supported by our study in this

White British sample. That is, the LOT-R measured two negatively correlated and

independent constructs. This result was consistent with previously reported analysis

in Chapter 2.

Testing for measurement invariance of LOT-R across cultures

A two-factor model of dispositional optimism was supported in previous SEM

analysis in Chapter 2.2.2. That is, the LOT-R measures two negatively correlated and

independent constructs. Similarly, a two-factor model of dispositional optimism was

supported in the White British sample as well. To test measurement invariance across

cultures, multi-group SEM was conducted. In addition to unconstrained base model,

Measurement weights, Structural covariances, and Measurement residuals were used

as constrained conditions in multi group analysis.

Fit statistics of all models tested are laid out in Table 5.3. Table 5.3 shows

that the unconstrained model fits best for the data. Among three constrained models,

Measurement weights model and Structural covariances model have similar fits as

the unconstrained baseline model. However, the absolute CFI value between

Measurement residual model and the unconstrained model is bigger than .05. It

means that for the unconstrained model, Measurement weights model and Structural

covariances model, ASQ-Positive structure is identical in measuring attributional

style across two cultures. However, for Measurement residual model, ASQ-Positive

structure doesn't have cross-culture validity.



Model NFI

Delta1

RFI

rho1

IFI

Delta2

TLI

rho2 CFI △CFI

Unconstrained .915 .841 .955 .912 .953




Table 5.3: Baseline comparisons for tested LOT-R models

Correlations between dispositional optimism, dispositional pessimism and

explanatory styles in Mainland Chinese and White British groups

Correlations for all the measures are presented in Table 5.4 for Mainland Chinese

(outside of parentheses) and White British (inside parentheses). As the table shows,

the pattern and magnitude of associations between measures for Mainland Chinese

and White British groups were quite similar. For example, dispositional optimism

scores were positively and significantly correlated with ASQ Total scores for both

Mainland Chinese (r = 0.13) and for White British (r = 0.17) groups; LOT-R

Pessimism scores were negatively and significantly associated with ASQ Positive for

both Mainland Chinese and for White British participants at the same level (r = -

0.23).

However, of the 21 pairs of correlations between the two cultural groups, we

still found some different patterns of correlations. For example, significantly weaker

negative associations emerged for White British participants compared with their

Mainland Chinese counterparts between dispositional optimism and dispositional

pessimism (r = -0.16 vs. r = -0.22, respectively), and between hopelessness and ASQ

Total scores (r = -0.63 vs. r = -0.54 respectively). More strikingly, while the

association between LOT-R Pessimism and Hopefulness scores was positive for

Mainland Chinese (r = 0.08), it was negative for White British (r = -0.05)



participants. Though neither of these correlations reached statistical significance,

they partly represented different trends of associations between explanatory style and

dispositional pessimism for these two ethnic groups. As a result, the association

patterns between these study variables was not identical for Mainland Chinese and

White British participants.

Measures ASQ-

Negative

ASQ-

Positive

ASQ

Total Hopelessness Hopefulness

LOT-R

Optimism

ASQ Negative -

ASQ Positive 0.28**

- (0.31**)

ASQ Total -0.63** 0.57**

- (-0.55**) (0.62**)

Hopelessness 0.95** 0.23** -0.63**

- (0.91**) (0.23**) (-0.54**)

Hopefulness 0.34** 0.94** 0.46** 0.32**

- (0.34**) (0.91**) (0.52**) (0.33**)

LOT-R Optimism -0.04 0.12 0.13* -0.04 0.11

- (-0.09) -0.11 (0.17*) (-0.11) -0.08

LOT-R Pessimism -0.07 -0.23** -0.25** 0.08 -0.17** -0.22**

(-0.02) (-0.23**) (-0.18**) (-0.05) (-0.16*) (-0.16*)

Table 5.4: Correlations for all measures


of parentheses are for White British. Correlations outside parentheses are for

Mainland Chinese. Hopelessness = stability + globality of the ASQ negative events;

Hopefulness = stability + globality of the ASQ positive events.

* p < 0.05. ** p < 0.01.

Cultural differences in dispositional optimism, dispositional pessimism,

attributional styles, and self-serving attributional bias between Easterners and

Westerners

Table 5.5 presents the results of t-tests comparing differences in dispositional

optimism, dispositional pessimism, Composite Negative Attributional Style (ASQ



Negative), Composite Positive Attributional Style (ASQ Positive), Composite

Positive minus Composite Negative (ASQ Total), Internal Negative, Stable Negative,

Global Negative, Internal Positive, Stable Positive, Global Positive, hopefulness, and

hopelessness. There were 13 planned comparisons assessing differences between the

two ethnic groups.

Measures

Culture group

t （432） Mainland Chinese White British

Means (SD) Means (SD)

ASQ Total 2.12 (2.24) 1.26 (2.21) 4.02***

ASQ Negative 12.98 (1.92) 12.21 (1.83) 4.26***

ASQ Internal Negative 4.47 (0.61) 4.34 (0.78) 2.06**

ASQ Stable Negative 4.33 (0.89) 4.03 (0.83) 3.60***

ASQ Global Negative 4.18 (0.96) 3.84 (0.89) 3.72***

Hopelessness 4.25 (0.81) 3.94 (0.76) 4.18***

ASQ Positive 15.1 (1.81) 13.47 (1.94) 9.05***

ASQ Internal Positive 4.87 (0.69) 4.58 (0.88) 3.89***

ASQ Stable Positive 5.29 (0.78) 4.63 (0.82) 8.57***

ASQ Global Positive 4.94 (0.80) 4.26 (0.79) 8.82***

Hopefulness 5.12 (0.69) 4.45 (0.70) 9.97***

LOT-R Optimism 8.37 (1.93) 7.02 (2.38) 6.50***

LOT-R Pessimism 4.05 (2.23) 4.51 (2.15) - 2.19**

Table 5.5: t-tests of ASQ and LOT-R between two cultural groups.


of parentheses are for White British. Hopelessness = stability + globality of the ASQ

negative events; Hopefulness = stability + globality of the ASQ positive events.

** p < 0.01. *** p < 0.001.



As shown in Table 5.5, Mainland Chinese participants reported significantly

higher dispositional optimism scores than White British (M = 8.37 vs. M = 7.02,

respectively), and significantly lower dispositional pessimism scores (M = 4.05 vs.

M = 4.51, respectively). The former result was quite unexpected given previous

findings obtained between Easterners and Westerners (Chang, 1996). But the

difference of pessimism scores was consistent with at least one study (Chang et al.,

2003).

Also as Table 5.5 shows, Mainland Chinese participants reported

significantly higher ASQ Negative scores than White British participants (M = 12.98

vs. M = 12.21, respectively), indicating a more pessimistic explanatory style for

negative events, which was consistent with previous findings (Lee & Seligman,

1997). At the same time, however, Mainland Chinese participants reported

significantly higher ASQ Positive scores than White British participants (M = 15.10

vs. M = 13.47, respectively), indicating that Mainland Chinese participants had a

more optimistic explanatory style for positive events than White British participants.

This result seemed quite unexpected given most previous findings obtained with

Asians and North Americans (e.g. Lee & Seligman, 1997), but it was consistent with

our previous findings that individuals tend to have a similar cognitive style for both

positive and negative events (see 2.1 in Chapter 2 for details). That is, people are

inclined to explain life events using consistent cognitive style, such as attributing

both positive and negative events to internal factors.

In spite of the difference of explanatory styles described above between these

two culture groups, both ethnic groups reported that ASQ Total scores were above

zero, indicating higher scores on positive events than on negative events (see Table

5.5). These results were consistent with previous findings reported by Higgins and

Bhatt (2001).

To further investigate potential cultural differences in explanatory styles

between these groups, t-tests were also conducted based on each of 12 ASQ events.

As shown in Table 5.6, Mainland Chinese participants reported higher scores on all

12 ASQ events than White British participants, indicating a more optimistic



attributional style for positive events and a more pessimistic attributional style for

negative events, which once again was consistent with the previous proposal of a

compatible cognitive style in explaining life events.

Life events Mainland

Chinese

White

British

Mean

Difference

Positive events Mean SD Mean SD

Achievement Becoming very rich 15.39 3.04 13.94 3.14 1.45 ***

Getting a position that you want very badly 16.00 2.71 13.72 2.89 2.28 ***

Getting a raise 15.21 2.49 13.65 2.88 1.56 ***

Affiliation Being complimented on appearance 13.63 2.84 11.84 2.88 1.79 ***

Being praised for doing a project 15.34 2.71 13.67 2.91 1.68 ***

Being treated more lovingly 15.03 2.95 14 3.24 1.03 **

Negative events

Achievement Having been failed to get a job for some time 13.35 2.74 12.51 3.1 0.84 **

An important talk gets negative reactions 13.25 3.02 12.61 2.68 0.64 *

Cannot meet expectations of others 13.53 2.81 12.42 2.58 1.12 ***

Affiliation Not helping a friend who has a problem 12.47 3.32 11.63 2.98 0.83 **

Being treated hostilely by a friend 12.61 2.92 12.03 2.47 0.58 *

A date goes badly 12.65 2.84 12.04 2.63 0.61 *

Table 5.6: Mean scores of Negative and Positive Affiliation and Achievement event

in two groups.

Note: For Mainland Chinese, N=232. For White British, N=202.

* p < 0.05.** p < 0.01. *** p < 0.001.



5.3.3 Are Chinese people more optimistic than British people?

The main purposes of the present study were fourfold. The first aim was to test

whether the same psychometric structures of the ASQ (three-correlated-factor

structure) and the LOT-R (two-factor structure) discussed in Chapter 2 were

replicable in a White British sample. I found that a model of causal attributions in

terms of three correlated factors of globality, stability, and internality adequately

accounts for responses to positive ASQ events but not for negative events. For

dispositional optimism, just as reported previously in most studies and in the SEM

analysis in Chapter 2, a two-factor model of dispositional optimism was supported in

this White British sample. That is, the LOT-R measured two negatively correlated

and independent constructs.

Second, in my attempts to find potential differences in optimism correlations

between two ethnic groups, the overall results revealed several critical points. First,

the patterns of associations between optimism measures for Mainland Chinese and

White British participants were quite similar. Fifteen out of twenty-one correlations

were found to be statistically significant for both groups (see Table 6.2). In sum,

ASQ Total was negatively correlated with LOT-R Pessimism and positively

correlated with LOT-R Optimism. As expected, LOT-R Optimism and LOT-R

Pessimism was negatively correlated. However, correlational patterns between

measured variables were not identical for two cultural groups (such as a weaker

negative association between LOT-optimism and LOT-pessimism for White British

participants than for Mainland Chinese participants).

Finally, in attempting to examine potential group differences on dispositional

optimism and explanatory style, I found that Mainland Chinese and White British

students differ among a number of important outcome variables in optimism.

Specifically, Mainland Chinese participants were significantly more optimistic and

less pessimistic. Also, Mainland Chinese participants showed a more pessimistic

explanatory style for explaining ASQ negative events than did their White British

counterparts, which supported the proposal that Easterners tend to use more

pessimistic attributions for negative events than Westerners. On the other hand,



although the difference in explaining ASQ positive events indicated a more

optimistic attributional style for Mainland Chinese participants, which was

seemingly inconsistent with some previous research, it supported the assumption that

individuals tend to produce similar patterns of explanations based on cognitive style

rather than on event type. Generally, these mixed results suggested that the cultural

influence on optimism is not uniform for at least some of the differentiated

dimensions.

The present findings demonstrate a trend of reversing traditional

understanding in assuming that Easterners are basically more pessimistic than

Westerners and Westerners are generally more optimistic than Easterners. These

findings appear inconsistent with many previous studies in which greater pessimism

was found in Easterners than Westerners. For example, Heine and Lehman (1995)

reported that the Japanese sample were more pessimistic than their Canadian

counterparts. Similarly, Lee and Seligman (1997) have also pointed to the greater

pessimism of Asians compared to European Americans. Therefore, we didn't expect

the opposite results. In spite of that, a few considerations may be helpful to account

for the lower pessimism found among Mainland Chinese compared to White British.

First, it has been argued that broader social factors should be taken into account in

understanding optimism and pessimism (Lee & Seligman, 1997). Accordingly, these

seemingly unexpected findings might be unique to this young Chinese population.

The relatively recent fast economic growth of China may provide an explanation for

Chinese people, especially as young generations feel more optimistic and confident

than previously, therefore dimming previous cultural influences on optimism.

Secondly, as noted by some researchers, one of the major concerns in

examing culture differences in optimism is that it might be a problem for Easterners

to get the exact meaning of LOT-R items since this questionnaire has been developed

on the basis of Western cultures (Anderson, 1999). Hence, it is possible that there are

slight gaps in understanding the meaning of optimism and pessimism. At the very

least, this is in line with some results from previous research, as discussed earlier,

that found no group differences in optimism across cultures (Chang et al., 2003), or

differences that were more nuanced (Chang, 1996).



Finally, in spite of differences in explanatory style between these two cultural

groups, the universality of the self-serving bias in causal explanations was supported

by the data. Both these ethnic groups reported positive ASQ Total scores, indicating

no matter what their cultural background was, individuals tend to explain positive

events with more internal, stable and global causes than negative events. This

conclusion is consistent with previous cross-cultural evidence (e.g., Higgins & Bhatt,

2001), revealing that there is a universal trend of positive bias in causal attributions.

We’d better bear in mind that though some specific patterns of optimism expression

are carved with potential cultural difference, it is generally true that being optimistic

means better psychological adjustment and is associated with higher levels of

happiness.


Chapter 6: Extending thoughts on attributional bias 158

Chapter 6: Extending thoughts on attributional bias

6.1 What we know and what we don’t know about attributional bias

One of the prevailing ideas in psychology is that individuals have an inherent and

pervasive tendency to provide explanations for the behaviour and events that they

encounter (Peterson, 2000a). As one of the most important psychosocial systems of

optimism, attributional style has been in attention of a large body of research, which

provides consistent evidence for the linkage between attributional style and many

other psychological traits. Such attributions can be functional and adaptive and may

serve psychological and social purposes when attributional bias applies (Mezulis et

al., 2004; Sanjuan & Magallares, 2014).

Attributional bias is argued to manifest itself in two related but distinct forms.

One is self-serving attributional bias (Mezulis et al., 2004). This refers to the

tendency of individuals to attribute positive situations to causes that are more internal,

stable and global than to causes for negative situations. The second form is self-

versus-other attributional bias – the tendency of individuals to attribute their own

behaviours to situational or environmental causes, while attributing behaviours of

others to dispositional or inherent causes (Ashkanasy, 1997). The literature focusing

on these two attributional biases are reviewed below.

Self-serving attributional bias

The original theoretical basis of self-serving attributional bias was that it derives

from the interaction between motivation and cognition certainty, suggesting that

people tend to “accept responsibility for positive behavioural outcomes and to deny

responsibility for negative behavioural outcomes” (Bradley, 1978, p. 59). Prior

studies addressing self-serving attributional bias used to focus solely on the

dimension of internality by assuming that individuals exhibit more internal

attributions for positive events than for negative events (Greenberg et al., 1982;

Nurmi, 1992). This concept was broadened by two facts. One is the development of a

widely-accepted three-dimensional measure for attributions – the ASQ. The other is



the rising debate of insufficient information for establishing a self-serving pattern in

attributions based only on the internality dimension. Consequently, the dimensions of

stability and globality have been incorporated, and self-serving attributional bias is

conceptualized as the tendency of people to attribute positive situations to more

internal, stable and global causes than they do for negative situations (Mezulis et al.,

2004).

Past studies have linked self-serving attributional bias to different aspects of

well-being. Sanjuan and Magallares (2014) reported positive relations between self-

serving attributional bias and two significant markers of well-being, subjective well-

being (r = .35) and adaptive coping strategies (r = .31). One of the earlier studies

found that depressed individuals were immune from self-serving attributional bias

while non-depressed subjects expressed apparent self-bias in causal attributions

(Greenberg, Pyszczynski, Burling, & Tibbs, 1992). Self-serving attributional bias has

also been implicated in the decision-making process, indicating that the preference of

attributing positive performance to internal causes increases confidence of financial

managers, and thus improve future performance as a result (Libby & Rennekamp,

2012).

In addition to research interested in the adaptive nature of self-serving

attributional bias in promoting well-being, psychologists have also investigated

potential influences of age, gender, and culture on this bias (Higgins & Bhatt, 2001;

Mezulis et al., 2004; Nurmi, 1992). Findings of these studies were basically

consistent with traditional understanding of culture differences between the East and

the West.

Though it is still not very clear what the inherent cognitive mechanism of self-

serving attributional bias is, evidence from an fMRI study has identified that this

type of bias is correlated with activation of the anterior portion of the precuneus

(Cabanis et al., 2013). This finding provides evidence for the physiological basis of

self-serving attributional bias.



Self-serving attributional bias and optimistic explanatory style

Comparing the definitions of optimistic explanatory style and self-serving

attributional bias, it is not difficult to see that both concepts share a favourable

attributional style involving both negative and positive situations. Similarity between

these two notions is strengthened by their methods of measurement. While a more

optimistic attributional style for a domain means higher scores for positive events

and a lower score for negative events for that domain (Forgeard & Seligman, 2012),

a self-serving attributional bias represents a positive score when attributions for

negative outcomes are subtracted from attributions for positive outcomes (Sanjuan &

Magallares, 2014).

Self-serving attributional bias in most current studies represents the positive

tendency in people’s causal attributions, and refers to an optimistic explanatory style,

which shows a cognitive bias in preference of an optimistic explanatory style, and

reflects a broad self-serving bias in attribution.

Prior research along both lines of optimistic explanatory style and self-serving

attributional bias are consistent in their finding of beneficial effects on well-being

(Forgeard & Seligman, 2012; Mezulis et al., 2004). For reasons of consistency, here

in my study of positive bias in attributions, the tendency of holding an optimistic

explanatory style and the tendency of expressing a self-serving attributional bias are

equal notions, both referring to the tendency for individuals to explain positive

situations through internal, stable and global causes, and negative situations to

external, unstable and specific causes.

Reflected in the ASQ, two composite scores, the ASQ Negative and the ASQ

Positive, were used to calculate a self-serving attributional bias (Sanjuan &

Magallares, 2014) or an optimistic explanatory style (Peterson et al., 1982). If the

subtraction score of the ASQ Negative from the ASQ Positive is positive, it

represents a self-serving attributional bias or an optimistic explanatory style,

reflecting stronger attributions along internal, stable and global causes for positive

than for negative events. On the other hand, if the subtraction score of the ASQ



Negative from the ASQ Positive is negative, it then stands for the lack of a self-

serving attributional bias or an optimistic explanatory style, reflecting weaker

attributions for positive than for negative events.

Self-versus-other bias in attribution of causality

Self-versus-other bias in attributions emerges when individuals attribute their own

performance outcomes to situational factors, and attribute others’ performance

outcomes to dispositional or internal factors (Ashkanasy, 1997). This notion of self-

versus-other attributional bias was originally developed based on Jones and Nisbett

(1972)’s proposition of actor-observer discrepancies or the actor-observer asymmetry.

Jones and Nisbett (1972) proposed in their theoretical analyses that based on

differences of information available for decision-making and different perspectives

on understanding personality of self and of others, individuals tend to attribute their

own behaviours to situational or environmental cause while attribute dispositional or

inherent causes for behaviours of others. This self-versus-other bias in attributions of

causality has become a common research topic in both psychology and sociology

(see Ashkanasy, 1997; Malle, 2006; Medway & Lowe, 1976; Teglasi & Fagin, 1984;

Watson, 1982). It has been connected to many potentially influential factors, such as

achievement (Medway & Lowe, 1976), social anxiety (Teglasi & Fagin, 1984),

psychosis (Wiffen et al., 2013), and perception of others (Ashkanasy, 1997).

The self-other view might also be viewed as an application of the self-serving

attributional bias, assuming that people tend to attribute their own success using

more internal causes than others’ success, and explain their own failure more

externally than others’ failure (Ashkanasy, 1997). Similar to assessment of self-

serving bias, the method of providing explanations for positive and negative

outcomes has been used widely in assessing self-versus-other attributional bias

(Malle, 2006). The outcome valence (positive-negative) has been taken as one of the

moderators of the self-other bias: Malle (2006) reported in his meta-analysis that the

self-other biased view is detectable in the case of explaining negative events but not




The moderating effect of interpersonal perception of the other has been

investigated. For example, Ashkanasy (1997) reported that when another individual

was seen to be similar to self, participants gave more internal causes to academic

success for others than they did for themselves, and gave more external causes to

academic failure for others than they did for themselves.

Though theory of self-versus-other bias in causal attributions has been developed

and assessed in some studies, there is no widely accepted definition and measure so

far since specific measurement for situational and dispositional causes haven’t been

developed.

6.2 Attributional evaluation system and possible attributional models

Attributional evaluation system and attributional models

If we are to understand the mechanism of attributional features, and to systematically

evaluate the potential relationship between two forms of attributional bias, it is

important that we systemically consider all components in the complex admixture of

attributions including subjects (self vs other), valences (positive vs negative events),

and causes (traits vs states) (see Table 6.1). Here, traits refer to inherent or fixed

aspects of causal attributions – internality, stability, and globality. Additionally,

states mean external or changeable features of attributions, representing the

dimensions of externality, instability, and locality. The possibility of modelling self-

serving bias and self-other bias in causal attributions jointly raises the possibility of

addressing the question whether attributions regarding the causes of positive and

negative events could be differentiated between self and other, i.e., do individuals

give more optimistic explanations for themselves than for others when both positive

and negative events apply?

Although theoretically positive or negative events could be attributed to either

traits or states independently, at least two extreme attributional styles, one of which

features attributing both good and bad situations to traits (see Table 6.1; system 1 and



system 5), and the other attributing both positive and negative outcomes to states (see

Table 6.1; system 4 and system 8), could be plausibly excluded. Moreover, previous

research has tested and confirmed self-serving attributional bias. As a result, my

understanding of causal attributions of self has predominately focused on models of

attributing positive events to traits of self, and attributing negative events to states of

self (see Table 6.1; system 3). For self-other attributional bias, based on previous

evidence of self-other attributional bias in at least the internality dimension

(Ashkanasy, 1997), we predicted that individuals would provide more biased

attributions for their own situations than they do for those of others.

Subject Valence

Attributions for Positive events Attributions for Negative events

Self

System 1 traits of self traits of self

System 2 states of self traits of self

System 3 traits of self states of self

System 4 states of self states of self

Other

System 5 traits of other traits of other

System 6 states of other traits of other

System 7 traits of other states of other

System 8 states of other states of other

Table 6.1: Computational structure of the attributional evaluation systems.

Thus, two attributional models were created to describe potentially true

evaluation patterns of causal attributions on the basis of analysis of the attributional

evaluation systems. The first model combines attributional evaluation system 3 and

system 6 (see Table 6.1); featuring two entirely opposite attributional styles between

self and other (Model A, see Figure 6.1). In this reversed model, individuals attribute

their own positive events to traits of self, and attribute other’s positive events to

states of other people; simultaneously, individuals tend to attribute their own

negative events to states of self, and attribute other’s negative events to traits of other



people. The second model stands for similar attributional patterns between self and

other (see Table 6.1, system 3 and system 7), but also features biased self-other

attributions. In this model, in addition to self-other discrepancy in causal attributions,

individuals are supposed to apply similar trends of optimistically-biased attributions

no matter what events occur to themselves or to other people (Model B, see Figure

6.2). That is, individuals tend to attribute positive events to traits and attribute

negative events to states both for themselves and for other people, though they tend

to give more credit for attributing their own behaviours.

Figure 6.1: Model A – reversed attributional model for self and for other.



Figure 6.2: Model B – more optimistic attributional model for self than for other.

Both models may reveal the truth, indicating that individuals tend to attribute

their own positive situations to more internal, stable and global causes than they did

for others in the same situations, while they tend to attribute more external, unstable

and local causes to themselves than they do for other people when negative situations

apply. Our aim was to test which model was the best attributional model when

individuals were asked to attribute the same events to themselves and other people.

Measuring issues

To investigate the possible attributional style in perception of others, we needed to

instruct participants to give attributions for themselves and others based on the same

events. So we administered a rewritten version of the ASQ, the ASQ-Other, asking

subjects what attributions they would make should these events occur to a fictional

character “Wang Chen”. Here “Wang Chen” is described as being a healthy

undergraduate with average intelligence. Subjects were asked to imagine each of a



series of events occurring to “Wang Chen”. The same 12 events were used as the

original ASQ.

6.3 Psychometric structure of the ASQ-Other

Before comparing causal attributions for the self and for the other, we first

investigated the psychometric structure of the ASQ-Other.

Participants in sample 1 (N = 452; for details, see 1.5.4 of Chapter 1) were

instructed to complete the ASQ-Other.

Analysis strategy

Descriptive statistics and correlational analyses were calculated first. Structural

equation modelling (SEM) was then used to test potential structural models of the

ASQ-Other using Amos 17.0 (Arbuckle, 2008). All analyses took advantage of raw

data supporting estimation of models using full information maximum likelihood

estimation.



explanatory style scores. Table 6.2 shows the descriptive statistics. Reliabilities were

acceptable.












ASQ-Other Total 0.96 1.27 0.89

Table 6.2: Means, standard deviations and Cronbach’s Alpha for all measures of the

ASQ-Other. (n = 452)



Modelling

The hypothesised three-factor model for negative events was tested using an MTMM

structure. The base model fitted reasonably well (χ² (114) = 225.59, p < .001; CFI =

0.94; TLI = 0.92; AIC = 339.59; BIC = 344.61; RMSEA = 0.047), but modifications

were suggested. The resultant model was a good fit by all criteria (χ² (109) = 168.58,

p <.001; CFI = 0.97; TLI = 0.96; AIC = 292.58; BIC = 547.63; RMSEA = 0.033), as

shown in Figure 2.4. Thus, as reported in the ASQ model earlier, a model of causal

attributions for negative events in terms of three correlated factors of globality,

stability, and internality adequately accounted for responses to these events in the

ASQ-Other as well. In this correlated factor model, stability and globality

correlated .58, internality and globality had an r of .27, and internality and stability

factors correlated .23.



Figure 6.3: Well-fitting 3-factor model of attributional style for others for negative

events.

Thus, a model of causal attributions for others for negative events in terms of

three correlated factors of globality, stability, and internality adequately accounted

for responses to these negative events in the ASQ-Other. This three-correlated-factor

model is also applicable in attributions of negative events when considering another

person being in the same situation, compared to attributions made when considering

the self in that situation.

.38

.50 .29 .15 .32

STABLE

.64 .66 .64

.63 .68 .60

GLOBAL

.55 8

.64 7

.69 .47 .65 .64

.13

.58

.27

.11

.08

.22

.26

.22

.

11

INTERNAL

.35

.31

.32

.13

.17

.22

.24

.31

.02

-.09

.03

.20

.02

.19

.20

-.19

.20

.11

-.10

2

4

5

7

8

5

2

4

7

8

11

2

4

5

11



A model for positive events was constructed in the same fashion as the baseline

model for negative events using the same MTMM structure (see Figure 2.5). Fitted

measures for the base model indicated adequate fit between model and data (χ2

(114)

= 239.21, p < 0.001; CFI = 0.94; TLI = 0.93; AIC = 353.21; BIC = 587.69; RMSEA

= .049), but modifications were suggested. The resultant model was a good fit by all

criteria (χ² (109) = 185.48, p <.001; CFI = 0.97; TLI = 0.95; AIC = 309.48; BIC =

564.53; RMSEA = 0.039). In the correlated factor model stability and globality

correlated .65, internality and globality .26 and internality and stability .43,

considerably higher than was the case for negative events.

Analyses of ASQ-Other positive and of ASQ-Other negative events, then,

indicated that these scales were well accounted for by three correlated factors of

internality, stability, and globality. That is, attributions regarding events that

occurred to others were well accounted for by the same three-correlated-factor

structure as the attributional style for explaining events occurred to self.



Figure 6.4 Well-fitting 3-factor model of attributional style for others for positive

events

INTERNAL

STABLE

GLABAL

.62 .27

.34

.48 .45

1 .24

.72

.53

.67

.62

.71 .50

.64 .62 .68

.54 .67 .55

.43

.65

.26

.14

.15

.44

.19

.28

.34

.36

.25

.41

.35

.30

.27

-.04

.09

.31

.15

.21

.13

.17

.18

.15 -.21

.17

3

6

9

10

12

1

3

6

9

10

12

1

3

6

9

10

12



6.4 Study 1: testing attributional models using ASQ and ASQ-Other

Participants in sample 1 were instructed to complete the ASQ and the ASQ-Other (N

= 452; for details, see 1.5.4 of Chapter 1).

Measures

Attributional style was assessed using the Chinese ASQ (Zhang, 2006). Attributional

style for others was measured using the ASQ-Other.

Procedure

Participants were tested in groups of 30 to 50 by their teacher. Each teacher was


participants completed the paper-and-pencil questionnaires. For the ASQ,

participants were instructed to make causal attributions for each of the 12 events

based on imaging that it occurs to them in real life. For the ASQ-Other, students

were asked to give explanations for the same life event when it occurred to other

people. Testing took around 30 minutes in total.

Scoring

Calculation of self-serving attributional bias followed the assessment method used in

Sanjuan and Magallares (2014).

Calculation of self-versus-other bias in attributions adapted a similar

assessment method to the ASQ. Specifically, if the subtraction score of the ASQ

positive from the ASQ-Other Positive is positive, it represents a self-other

attributional bias, reflecting stronger attributions along internal, stable and global

causes for self than for other for the same positive events. On the other hand, if the

subtraction score of the ASQ-Other Negative from the ASQ negative is positive, it

also stands for a self-other attributional bias, revealing a more optimistic explanatory

style for self than for other for the same negative events.



Results

Descriptive and summary statistics and the standard composite attributional style

scores of the ASQ are shown in Table 6.3. See Table 6.2 for the descriptive statistics

of the ASQ-Other for the total sample.










ASQ Total 2.38 2.17 0.84

Table 6.3: Means, SDs and Cronbach’s Alpha for the ASQ scales.

As shown in Table 6.3, all dimensions for ASQ positive events, including ASQ

Positive, Internal Positive, Stable Positive, and Global Positive, scored higher than

the four corresponding dimensions for negative events. As a result, the subtraction

score of the ASQ Negative from the ASQ Positive is positive. Similarly, as shown in

Table 6.2, all measuring dimensions for ASQ-Other positive events, including ASQ-

Other Positive, Internal Positive, Stable Positive, and Global Positive, scored higher

than four corresponding dimensions for negative events. As a result, the subtraction

score of the ASQ-Other Negative from the ASQ-Other Positive is positive.

In order to test self-other attributional bias, t-tests were conducted and mean

differences revealed that there were significant differences between scores of all

dimensions measured in the two questionnaires (see Table 6.4). The results show that



participants had significantly higher composite scores on positive events of ASQ

than on positive events of ASQ-Other, and participants scored significantly lower on

ASQ negative than they did on ASQ-Other negative.

Dimensions ASQ ASQ-Other

Means (S.D.) Means (S.D.)

Positive Events

Internal Positive 5.03 (0.70) *** > 4.45 (0.69)

Stable Positive 5.36 (0.78) *** > 5.33 (0.82)

Global Positive 4.90 (0.85) < 5.15 (0.84) ***

Total 15.28 (1.91) *** > 14.93 (1.81)

Negative events

Internal Negative 4.45 (0.67) *** > 4.04 (0.65)

Stable Negative 4.33 (0.85) < 5.10 (0.87) ***

Global Negative 4.12 (0.90) < 4.83 (0.89) ***

Total 12.90 (1.78) < 14.00 (1.74) ***

Table 6.4: t-tests between ASQ and ASQ-Other for attributional style.

*** p < 0.001.

Mixed results emerged with regard to specific dimensions of the ASQ and ASQ-

Other. For positive events, participants scored significantly higher on internality and

stability but significantly lower on globality of the ASQ than they did on

corresponding dimensions of ASQ-Other; for negative events, participants scored

significantly lower on stability and globality but higher on internality of the ASQ

than they did on corresponding dimensions of ASQ-Other.

This self-other discrepancy of causal attributions was also generalized along

with three attributional dimensions: for Internality, participants showed significantly

higher scores for both ASQ positive and negative events than they did for ASQ-

Other positive and negative events; for Stability, subjects reported significantly

higher ratings for ASQ positive events than they did for ASQ-Other positive events



but significantly lower ratings for ASQ negative events than for ASQ-Other negative

events; for Globality, participants scored significantly lower for ASQ positive events

than for ASQ-Other positive events but significantly higher for ASQ negative events

than for ASQ-Other negative events.

Finally, self-serving attributional bias and self-other attributional bias were

combined (see Figure 6.5). Participants scored higher in attributions for positive

events than for negative events when these events occurred to themselves, and they

scored significantly lower in attributions for negative events than they did for other

people for the same events.

Figure 6.5: Attributions for positive and for negative events, for self and for other.

Discussion

As expected, results indicated that positive self-serving bias was displayed in each of

the three attributional dimensions across event valence. When individuals attribute

causal explanations for life events, they prefer to give more internal, stable and

global causes for positive events than they do for negative events. For negative

12.50

13.00

13.50

14.00

14.50

15.00

15.50

Positive Events Negative Events

Self (ASQ)

Other (ASQ-Other)



situations, individuals have the tendency to attribute those situations to more external,

unstable and specific causes than they do for positive events. A self-serving

attributional bias is manifested in the ASQ, reflecting optimistically biased

attributions with internal, stable and global causes.

Turning to the hypothesis that the subject would show a self-other attributional

bias, results indicated that individuals tend to have a more optimistic explanatory

style for similar situations with themselves than with other people for both positive

and negative events. That is, people tend to explain events in their own best interest.

While people explain their own positive outcomes using more favourable internal

causes, they attribute others’ positive outcomes to external variables. Similarly,

people also tend to see their own negative situations to be externally caused than

others.

However, caution should be taken when applying this tendency for specific

dimensions of attributional style. Though generally ASQ Positive scores were higher

than ASQ-Other positive scores, which was also the case for dimensions of

internality and stability, participants scored lower on globality of the ASQ than they

did on corresponding dimensions of ASQ-Other. Similarly, participants scored

significantly lower on composite ASQ Negative than they did on ASQ-Other

Negative, which was also applicable for dimensions of stability and globality, but the

dimension of internality was not consistent with this trend. These two exemptions

have no much influences on the general conclusion that individuals show a self-other

bias in causal attributions, because we should bear in mind that it is recommended in

ASQ scoring that the composite scores (ASQ Positive and ASQ Negative) values

much more than individual dimension scores (Peterson et al., 1982).

Model B (see Figure 6.2), which represents similar attributional trends between

the self and the other but also features a more optimistic attributional style for the

self than for the other, was supported by the data. Individuals provide more

optimistic explanations for positive outcomes than they do for negative events for

their own behaviours. At the same time, they hold a more optimistic explanatory

style when the same event is explained for themselves than for others, no matter if it



is for positive or negative events. Data analysis supported the validity of Model B,

though we found that there are bigger discrepancies between ASQ Negative and

ASQ-Other Negative than differences between ASQ Positive and ASQ-Other

Positive (see Figure 6.5). Why is there less discrimination among attribution scores

for positive events than for negative events? Peterson et al. (1982, p. 295) explained

it as “perhaps people make fewer distinctions among good events since they may not

spend as much time ruminating over them as they do over bad events, and may

attend more to the causes of bad events”.

The results of this study suggest that attributions, whether for the self or for the

other, are optimistically biased. That is, individuals tend to attribute positive events

to inherent or fixed causes (traits) and attribute negative events to external or

changeable causes (states) both for themselves and for other people. One unanswered

question from this study is whether this optimistic bias holds equally for positive and

negative events, i.e., do we have a general tendency to be more optimistically biased

for attributing positive events than we are for attributing negative events? If this is

the case, then the next question is whether our attributions for self or for other people

are closer to this generally optimistically biased tendency. To address these questions,

we conducted a second study.



6.5 Study 2: testing event-focused attibutional style using ASQ-General

We have re-written the ASQ into a novel adapted version, the ASQ-General, asking

subjects what attributions they would make should these events occur to ‘someone’,

which could be themselves or any other person. Based on findings of a general

tendency of attributional biases in both the ASQ and the ASQ-Other in the first study,

it was predicted that this optimistically biased attributional style would also be

applicable in the ASQ-General. That is, when there are no specific subjects

designated to possible life events, individuals will tend to attribute positive situations

to causes that are more internal, stable and global than to causes for negative

situations.

Subjects

Participants in sample 4 were instructed to complete the ASQ-General (N = 117; for

details see 1.5.4 of Chapter 1).

Measure

The original ASQ is based on explanations for events (positive and negative)

imagined as occurring to the subject themselves. To investigate the possible

attributional style in general, the standard ASQ was modified as the ASQ-General,

asking subjects what attributions they would make should these events occur to

“someone” who represents not just the subject but all people. The same 12 events

were used as the standard ASQ: six positive (e.g. ‘someone does a project that is

highly praised’) and six negative (e.g. ‘someone has been looking for a job

unsuccessfully for some time’) events. Rating and scoring of the ASQ-General was

the same as the standard ASQ.



Procedure

Participants were tested in groups of around 30 by their teacher. Each teacher was


participants completed the paper-and-pencil questionnaires. Students were instructed

to “Write down one thing you think most commonly causes this situation (on average

for all people, not just you)”. Testing took around 20 minutes in total.

Analysis and results


attributional style scores. Table 6.5 shows the descriptive statistics of the ASQ-

General for the total sample. Reliabilities were acceptable.

In order to test attributional bias in general situations, t-tests were conducted and

mean differences revealed that there were significant differences among scores of all

the ASQ-General dimensions (see Table 6.6). The results showed that participants

had significantly higher composite scores on positive events than composite scores

on negative events, and had significantly higher scores on all three specific

dimensions of the ASQ-General.

Means SD Cronbach’s Alpha

ASQ-General Positive 14.96 1.91 0.82

ASQ-General Internal Positive 4.69 0.73 0.59

ASQ-General Stable Positive 5.17 0.85 0.76

ASQ-General Global Positive 5.1 0.84 0.71

ASQ-General Negative 13.92 2.08 0.83

ASQ-General Internal Negative 4.49 0.72 0.56

ASQ-General Stable Negative 4.75 0.86 0.73

ASQ-General Global Negative 4.68 0.98 0.78

ASQ-General Total 1.04 1.78 0.89

Table 6.5: Means, SDs and Cronbach’s Alpha for the ASQ-General scales.



Dimensions

Means (SD)


Internality 4.69 (0.73) *** > 4.49 (0.72)

Stability 5.17(0.85) *** > 4.75 (0.86)

Globality 5.10 (0.84) *** > 4.68 (0.98)

Total 14.96 (1.91) *** > 13.92 (2.08)

Table 6.6: t-tests between ASQ-General dimensions.

*** p < 0.001.

Putting the results of the two studies together (see Figure 6.6) showed that the

ASQ-Other and the ASQ-General almost overlap, while the ASQ was clearly

differentiated from the other two.

Figure 6.6: Attributions for positive and for negative events, for self, other and

general

12.50

13.00

13.50

14.00

14.50

15.00

15.50


Self (ASQ)

Other (ASQ-Other)

General (ASQ-General)



Discussion

Results from the second study showed higher composite scores on the ASQ-General

positive events than composite scores on ASQ-General negative events, as well as

higher scores on all three specific dimensions of the ASQ-General positive events

than for negative events. These results suggest that optimistically biased attributions

are also applicable in general situations. Individuals generated attributional style on

the basis of judgment for features of those events (positive or negative). No matter

whom was the subject experiencing these events, themselves or other people,

individuals showed a general attributional bias, indicating more internal, stable, and

global attributions for positive events than they did for negative events.

Comparing scores of the ASQ-General with the two measures in the first

study, we found that individuals show more positive bias towards themselves than

for other people or a general population in causal attributions, especially for negative

events.

6.6 Attributional biases in reality

There has been widespread recognition that attributional bias plays an important role

in the causal attributions that people make across event valence (positive vs negative

outcomes) and across perception (self vs other), and categorise them as self-serving

attributional bias and self-other attributional bias, respectively. Though these two

forms of attributional biases are theoretically connected (Ashkanasy, 1997), research

testing attributions for the causes of events occurring to others has been separated

from studies of attributional bias regarding the self, with no research including both

into a constructed evaluation system.

Prior research examining attributional bias has taken into account subjects (self

vs other), valences (positive vs negative events), or causes (traits vs states). Not all of

these components have been systemically reviewed in one single study. We

combined these critical components into the complex admixture of causal

attributions, generating eight potential attributional systems and two potential



attributional models. Using the most widely used assessment for causal attributions,

the ASQ, and a rewritten novel version of this instrument, the ASQ-Other, I first

tested which of the two models was the best attributional model when individuals

were asked to attribute the same events when they happened to themselves and to

other people.

Findings of the first study demonstrated that causal attributions about life events

possess self-protection features, as suggested by Heider (1958). Individuals tend to

maximise positive and minimise negative future outcomes in making attributions,

thus showing a self-protective bias in causal explanations for personal outcomes or

situations. As expected, I found that positive self-serving bias manifested in each of

the three attributional dimensions across event valence. When individuals attribute

causal explanations for life events, they prefer giving more internal, stable and global

causes for positive outcomes than for negative outcomes. For unfavourable situations,

individuals have the tendency of attributing those situations to external, unstable and

specific causes. Confirmation of self-serving attributional bias in this Eastern sample

provided further evidence to the universality of this positive bias (Mezulis et al.,

2004). It appears that there may be a universal tendency for individuals to protect

themselves against negative feelings by using an optimistic attributional style.

Regarding self-versus-other bias in attributions of causality, results supported the

idea that individuals do have biased attributions for what happens to themselves and

to others. This optimistically biased tendency applies to both positive and negative

events. While individuals attribute their own positive outcomes to dispositional

factors and attribute their own negative outcomes to situational factors, they tend to

attribute other peoples’ positive outcomes to situational factors and other people’s

negative outcomes to dispositional factors. As a result, in the two proposed potential

attributional models, Model B (see Figure 6.2) was supported with more biased

attributions for negative events than for positive events between the self and the other.

The first study suggests that attributions are optimistically biased for both the

self and the other. Individuals apply similar trends of optimistically biased

attributions no matter what events occur to themselves or to other people. This raised



the question whether this optimistic bias holds equally for positive and negative

events, i.e., are individuals more optimistically biased for attributing positive events

than they are for attributing negative events with a general tendency. This question

was tested in Study two using another rewritten version of the ASQ, the ASQ-

General. Results revealed that the optimistically-biased tendency in causal

attributions were generally applicable when there is no specific subject was

designated. People tend to attribute internal, stable, and global attributions for

positive events while they generate external, unstable, and specific explanations for

negative events no matter whether the subject is themselves or other people. In

summary, individuals generally show an optimistically biased attributional style

towards positive outcomes than they do for negative outcomes.

Previous studies examined either just one type of attributional bias or

investigated only the dimension of internality concerning self-other bias. My study

made it possible to combine self-serving bias and self-versus other bias in

attributions in a widely-accepted three-dimensional model of causal attributions. It

revealed that explanations for causes of positive events and negative events could be

differentiated between self and other. Individuals gave more optimistic explanations

for themselves than they did for others. This self-versus-other bias existed in

people’s attributions for both positive events and negative events.


Chapter 7: Depression, positive psychology and optimism interventions 184

Chapter 7: Depression, positive psychology and

optimism interventions

According to a report from the World Health Organization (2012), over 250 million

people are affected worldwide by depression, which is believed to lead to the suicide

of approximately 1 million people every year. Unfortunately, less than half of the

population affected by depression receive any effective physical or psychological

treatments. This figure is even less than 10 percent in some underdeveloped countries.

Insufficient information available for diagnosis can cause delays and improper

treatment for depression, and there is a lack of effective intervention resources that are

low cost and easily accessible (Sin et al., 2011).

Over the past 15 years, research in the field of positive psychology has shown

that psychological well-being can be cultivated and promoted through brief

interventions aimed at developing positive feelings, behaviours, or cognitions (Layous

et al., 2011; Seligman et al., 2006; Sin & Lyubomirsky, 2009). Diverse positive

psychology interventions have emerged and have provided empirical evidence for the

happiness-enhancing effect of individual strengths and resources. Unsurprisingly,

positive interventions can be particularly useful for the amelioration of depressive

symptoms, since depressed individuals will likely benefit from increases in positive

emotions (Sin et al., 2011).

Since optimism has been identified as having the strongest link to well-being in

the identified 24 character strengths in positive psychology (Park et al., 2004), and has

been shown to be beneficial in decreasing depressive symptoms (Sin et al., 2011), I

also wanted to look at the application of optimism interventions to depression

treatment, testing whether optimism manipulations could alleviate depressive

symptoms in the group of first-year college students. To understand mechanisms



underlying optimism interventions, I first reviewed several traditional treatments for

depression, and then turned to theoretical background and practical manipulations of

optimism interventions, which have been included into the increasing development of

positive psychology therapy.

7.1 Traditional treatments for depression

Currently, there are two main approaches to treating depression: physical and

psychological treatment. The main physical treatment is anti-depressant medication,

which addresses the neuro-transient of the chemical process underlying depression in

the brain. The molecular and biochemical origins of depression are still not fully

understood. It is not surprising, then, that current medication is suboptimal. For

example, for mild to moderate depression, there is no significant difference between

the effect of a treatment pill and a placebo, with more than 80% of the effect of the

anti-depressant drug accounted for by placebo effects (Kirsch, Moore, Scoboria, &

Nicholls, 2002). Another problem with anti-depressant treatment is the high risk of

relapse following the cessation of treatment (Layous et al., 2011)

There are a number of psychological treatments for depression that show

evidence of working well, such as Cognitive Behavioural Therapy and problem-

solving therapy. Cognitive Behavioural Therapy enables patients to correct false self-

beliefs that can lead to certain negative emotions and behaviours (Rupke, Blecke, &

Renfrow, 2006). American psychologist Aaron Beck is regarded as a pioneer in

cognitive therapy. Through his working with depressed patients, he found that

negative moods and behaviours were usually caused by distorted thoughts and beliefs

(Beck, 1976).



Three cognitive aspects – automatic thoughts, emotional responses, and

behavioural responses, have been identified as the cognitive view of human

functioning. It has long been debated that the spontaneous and immediate judgement

of a situation may be crucial in eliciting and shaping a person’s emotional and

behavioural responses to that situation. On the basis of this, Beck (1976) developed

the Cognitive Therapy (CT) for psychopathological treatment of depression. The

fundamental assumption behind CT is that a thought precedes a mood, and that both

thought and mood are interrelated with environment, physical reaction, and

subsequent behaviour. In this sense, the way people feel is related to the way in which

they explain and think about an event. The event itself does not directly determine

how they feel; their emotional response is mediated by their perception of the event (J.

S. Beck & Beck, 2011).

CT and interpersonal treatment have been shown to be effective for mild and

moderate depression. A meta-analysis of 15 studies on psychological treatments on

adult depression showed a standardised mean effect size of psychological treatment

versus control groups of 0.31 (Cuijpers, Van Straten, Van Schaik, & Andersson,

2009). Another more recent meta-analysis covering 1,036 studies on the effects of

psychotherapy for adult depression had a mean effect size of 0.42 after adjustment for

publication bias (Cuijpers, Smit, Bohlmeijer, Hollon, & Andersson, 2010).

Taken together, current medication treatments are criticised for their high

financial costs, potential side-effects, and limited effect (Layous et al., 2011). By

contrast, traditional psychological treatments have been shown to be effective in

reducing acute distress in depressed individuals and more preferable to drug therapy

among all but the most depressed people. However, these traditional psychological

treatments focus on alleviating depressive symptoms, and assume that mental health

equate to the absence of mental illness. This assumption makes traditional



psychological treatments vulnerable to newly-rising positive psychotherapy (Sin et al.,

2011). New treatments that can balance the advantages and deficits of medication

therapy and traditional psychology treatments are needed.

7.2 Rising of positive psychology interventions

According to the learned helplessness theory (Abramson et al., 1978) and its later

version, the hopelessness theory of depression (Abramson et al., 1989), depression is

conceptualized as an overabundance of negative moods and negative cognition. It is

the tendency to attribute internal, stable, and global causes to negative events that

results in hopelessness and thus depression. Depression treatments developed on the

basis of these ideas then predominantly focused on fixing and alleviating negative

feelings behaviours. Positive psychology grew from the recognition that a positive

state or trait is not necessarily the obverse of negative experiences and traits; and,

positive emotions and behaviours stand for a completely separate psychological

process that functions via an isolated neural mechanism (Duckworth et al., 2005).

If traditional depression treatment aims to cure mental illness by fixing

negative feelings and negative thoughts, positive psychotherapy strives to ameliorate

depressive symptoms by promoting positive affect and positive thoughts, such as

savouring (Bryant & Veroff, 2007), practicing forgiveness (Reed & Enright, 2006),

using signature strengths (Linley et al., 2010), and expressing optimism and gratitude

(Lyubomirsky et al., 2011). This has been shown to boost positive emotions, positive

thoughts, positive behaviours, and alleviating depressive symptoms (Layous et al.,

2011; Seligman et al., 2006; Sin & Lyubomirsky, 2009).

Positive psychology includes many traits that are associated with indices of

well-being. Twenty-four character strengths have been identified, in which optimism



was found to have the strongest link to life satisfaction – one of three significant

marks of well-being (Park et al., 2004). Additionally, numerous cross-sectional and

longitudinal studies have revealed that optimism is strongly correlated with a host of

psychological variables, such as self-esteem, academic achievement, coping strategy,

and positive emotions, and perhaps most importantly, predicts psychological and

physical well-being both in the presence and absence of stressors (Carver & Scheier,

2014; Carver et al., 2010; Forgeard & Seligman, 2012; Scheier & Carver, 1992).

Taken together, research suggests that optimism is associated with various indices of

positive functioning in a wide variety of stressful situations. To fully understand the

mechanisms underlying the beneficial effects of cultivating optimism in relationship

to depressive symptoms, I next turn to the theoretical background of the optimism-

depression relationship.

7.3 Optimism and depression

As stated in previous chapters, optimism has been conceptualized and measured in

different ways, among which dispositional optimism and optimistic explanatory style

are regarded as the two main contrasting approaches (Carver et al., 2010; Forgeard &

Seligman, 2012). No matter how optimism is conceptualized and measured, research

is uniform in indicating that optimism is bonded with beneficial characteristics:

happiness, achievement, health, and persistence. Considering all the direct and

indirect associations between optimism and personal and social benefits, it is not

surprising that optimism is reported to be relevant to clinical psychology. Results of

optimism interventions for depression have been both involved in the whole frame of

positive psychotherapy and taken as single treatment. The strength of optimism in

ameliorating depressive symptoms has received substantial empirical support (Csillik,

Aguerre, & Bay, 2012; Seligman et al., 2005; Sin et al., 2011).



7.3.1 Attributional style in depression

To understand presumptions behind the relationship between optimism and

depression and to find out the mechanisms under which optimism interventions works

for depression treatment, it’s necessary to first illustrate the theoretical assumptions of

related optimism theories.

Attributional models of depression

The causes and consequences of depression have long occupied the attention of

psychologists and clinical practitioners. Before the application of Seligman’s (1976)

learned helplessness model of depression, most theories and research had been

developed by clinical psychologists. Based on findings in psychological experiments

on animals, Maier and Seligman (1976) developed principles of “learned

helplessness”, assuming that helplessness occurs when there is an expectation of

uncontrollable events. In humans, only certain individuals respond pessimistically

after being exposed to uncontrollable aversive events.

To account for these findings, the learned helplessness model was refined into

the reformulated learned helplessness theory (Abramson et al., 1978), in which the

dimensions of attributional style – internal-external (Heider, 1958), stable-unstable

(Weiner, 1974), and global-specific (Abramson et al., 1978) (especially for negative

events) – were emphasised. An internal attribution explains the cause of a negative

event to factors inside the self, whereas an external attribution explains the cause in

self-referent terms. The more internal one’s attribution for lack of control is, the more

self-esteem will be lowered. A stable attribution assigns the causes of a negative event

with constant and perpetual factors across time, whereas an unstable attribution

explains the event in terms of momentary and time-limited factors.



Similarly, attributions may also vary in their degree of globality. A global

attribution assigns pervasive factors to causes of a negative event across different

situations, whereas a specific attribution explains a negative event in terms of

exceptional and situational factors. Accordingly, individuals who explain causes of

negative events with internal, stable, and global factors will be more vulnerable to

depression than those who provide attributions in terms of external, unstable, and

specific factors. Thus, the traditional study in depression was extended to the domains

of social and personality psychology, taking individual differences in attributional

style into account.


distinct behavioural responses. For instance, low self-esteem is agreed to be linked

with internal attributions regarding negative events, while chronic depression may

result from stable attributions for negative events (Haugen & Lund, 1998; Peterson et

al., 1982). In this learned helplessness model, depression emerges as a consequence of

experience with uncontrollable negative events (Abramson et al., 1978).

To expand earlier concepts, the hopelessness theory of depression was

developed from the reformulated learned helplessness theory. In addition to the

original presumption of helplessness, the expectation for the occurrence of negative

outcomes was added to construct the core concept of hopelessness. According to the

hopelessness theory of depression, hopelessness is conceptualized as the expectancy

that future outcomes will be stable, global, and will negatively influence many aspects

of an individual’s life regardless of his or her efforts (Abramson et al., 1989). As a

result, hopelessness about the future constitutes a sufficient and proximal cause of a

subtype of depression, called hopelessness depression (Abramson et al., 1989). This

attributional model of depression has accumulated substantial evidence from

empirical studies (e.g. Vazquez et al., 2001).



Though originally this depressive attributional style was applied mainly to

negative events, Seligman, Abramson, Semmel, and Von Baeyer (1979) suggested

that it might also play a part in explaining positive events. The authors found that

depressed students attributed good outcomes to more external and unstable factors

than did non-depressed students, and attributed more internal, stable, and global

causes to negative events than non-depressed students.

Studies on the attribution-depression relationship

Studies examining associations between attributional style and depression have been

conducted both from a cross-sectional perspective and a prospective approach,

involving adults, children, and adolescents. Cross-sectional studies propose that a

pessimistic attributional style is correlated with hopelessness and thus depression. On

the other hand, an optimistic explanatory style has been linked to protection from

depression. A pessimistic explanatory style predicts increases in depression over time

in different populations, such as lower-class women, children, and depressed patients

(Peterson & Seligman, 1984). Peterson and Vaidya (2001) reported that hopelessness

positively correlated with depression in their study with a group of college students (r

= .20).

In an earlier meta-analytic review, Sweeney, Anderson, and Bailey (1986)

reviewed 100 studies involving nearly 15,000 subjects. They found that attributions to

external, unstable, and specific causes for positive events and attributions to internal,

stable, and global factors for negative events were correlated with depression (average

r = -.15 and average r = .27 respectively). Haugen and Lund (1998) also reported a

negative correlation between ASQ Positive and depression (r = -.27), and a positive

correlation between ASQ Negative and depression (r = .20).



Subsequent studies have incorporated structural equation modelling (SEM),

allowing a better understanding of the relationship between attributional style and

depression by contrasting competing theoretical models. For instance, Ledrich and

Gana (2013) reported a SEM analysis of the attribution-depression relationship in 334

participants. EASQ was used to measure attributional style. The correlation between

pessimistic attributions for negative events and depression was .36. In addition to the

composite score, each of the three attributional dimensions, internality (r = .15),

stability (r = .19), and globality (r = .28) also positively correlated with depressive

mood.

Prospective studies collect longitudinal data to analyse the attribution-

depression relationship, which has been shown to be persistent over time (for a review,

see Wise & Rosqvist, 2006). For instance, Iacoviello, Alloy, Abramson, Whitehouse,

and Hogan (2006) examined whether cognitive style predicts the future development

of depression. One hundred and fifty-nine college students were divided into a high-

risk group and a low-risk group based on their scores of attributional style and

dysfunctional attitudes at baseline, and then were assessed for their depressive

symptoms every six weeks across a period of 2.5 years. This study showed that

cognitive high-risk participants experienced more episodes of depression, more severe

episodes, and more chronic courses than low-risk participants. The results suggested

that negative attributional style may confer risk for the development of depressive

symptoms.

Further, attributing life events along the dimension of globality may play a

significant part in predicting depression. For example, in a recent 10-month follow-up

study (n = 3500), Pearson et al. (2015) found that attributions to global factors for

negative events clearly correlated with future depressed mood in young adults. This



effect was independent of the other two dimensions of causal attribution, internality

and stability.

If it is true that a pessimistic attributional style interacts with adversity to

predict depression in the long run, does it mean that an optimistic explanatory style

interacting with positive events could reduce depressive symptoms? Haeffel and

Vargas (2011) tried to answer this question by asking 128 college students to

complete measures for depression, attibutional style (CSQ), and life events at baseline

and then reassessing them with the same questionnaires four weeks later. Results

indicated that participants with a pessimistic attributional style who experienced a

high ratio of stressful life events reported the greatest level of depressive symptoms.

However, they were buffered from depression and displayed similar levels of

depression with participants without a pessimistic explanatory style if they also

possessed an optimistic attributional style or had experienced many positive events.

These findings suggest that having an optimistic attributional style and experiencing

positive events may play a protective role against depressive symptoms.

Potential mediating roles of attributional style between depression and some

physical variables have been investigated. For instance, 23 depressed patients and 31

never-depressed controls completed the ASQ and a measure of sleep over a period of

seven days (P. L. Haynes, Ancoli-Israel, Walter, & McQuaid, 2012). Among the three

individual dimensions of attributional style, globality was found to mediate the

relationship between sleep disturbance (poor sleep continuity, delayed morning wake

time, and increased total time spent in bed) and depression.

The prospective relationship between attributional style and depression has

been reported in clinical settings as well. For instance, Sanjuán, Arranz, and Castro

(2012) conducted a two-wave longitudinal study in a group of 99 patients with

coronary heart disease. An adaption version of the original ASQ which contains only



six negative events was used to assess attributional style. The globality dimension was

associated with both Time 1 and Time 2 depressive symptoms (r = .26 and r = .34

respectively), while the stability dimension was only correlated with Time 2

depression (r = .20). For the dimension of internality, no significant correlations with

either Time 1 or Time 2 depressive symptoms were found. Additionally, global

attributions predicted persistence of depressive symptoms eight weeks later. These

results suggested that attributing negative events to pervasive and global causes lead

to increased depressive symptoms.

Using both a cross-sectional approach and a prospective design, Fresco, Alloy,

and Reilly-Harrington (2006) examined the relationship between causal attributions

and depression across a period of four weeks. Two hundred and thirty-nine

undergraduates were divided into either a currently depressed/anxious group or a

normal control group, and completed self-reported measures of attributional style,

depression, life events, and mood disorders, as well as structured diagnostic

interviews in two time slots. Results showed that participants in the depressed group

scored higher in attributions for positive events than their counterparts in control

group. Attributional style moderated the relationship between the occurrence of life

events and changes in depressive symptoms from Time 1 to Time 2.

Studies conducted in children and adolescents support the attribution-

depression relationship as well. For instance, 295 secondary school students were

instructed to complete measures of attributional style, self-esteem, and depression

(Kurtovic, 2012). This study indicated that attributing academic failure to stable and

global causes correlated with higher depression (r = .17 and r = .20 respectively),

while attributing academic success along stable dimension correlated with lower

levels of depression (r = .15). Additionally, hopelessness correlated significantly with

depression (r = .58). In a meta-analytic review of attribution-depression studies



conducted in children and adolescents (27 studies, 4,000 subjects), Joiner and Wagner

(1995) reported that attributional style scores clearly correlated with both self-

reported depression and with clinical depression (for overall composite scores,

average r = -.50; for positive events, average r = -.38; for negative events, average r

= .35).

7.3.2 Dispositional optimism and depression

Dispositional optimism has also been shown to be associated with depression.

According to the theory of dispositional optimism, being optimistic means having

favourable generalized expectations and continuing goal-pursuit for the future

(Scheier & Carver, 1993). Optimists expect good outcomes, which result in more

positive feelings and affections, while pessimists expect bad outcomes, and this yields

a relatively negative mix of feelings, such as anxiety, sadness, disappointment, and

anger (Scheier & Carver, 1992). Depression and distress sometimes occur due to these

negative feelings.

In a meta-analytic review of 56 studies (Andersson, 1996), the average

weighted correlation between dispositional optimism and depressive symptoms was -

.45. Peterson and Vaidya (2001) also reported that expectations (measured by the

LOT) were significantly correlated with depressive symptoms (r = -.55). Isaacowitz

(2005) addressed this issue in a wider range with three age groups (100 young, 86

middle-aged, and 94 older adults). The study reported that LOT optimism negatively

correlated with depressive symptoms across all three age groups (r = -.34, r = -.32,

and r = -.31 respectively), and LOT pessimism positively correlated with depression

in the middle-aged group (r = .29) and older adults group (r = .41). No significant

association between LOT pessimism and depressive symptoms of young adults was

found.



Stressful life changes may play a role in the relationship between dispositional

optimism and depression. One study examined the relationship between dispositional

optimism and depression in a small group of postnatal women (n = 75). The results

showed that LOT optimism was inversely correlated with depression both in initial

assessment (r = -.41) and three weeks later (r = -.43) (Carver & Gaines, 1987).

Armbruster, Pieper, Klotsche, and Hoyer (2015) examined whether

dispositional optimism reliably predicts depression across a period of five years.

Participants (n = 4,046) were divided into five age groups (18-44, 45-54, 55-64, 65-74,

and 75-84). They were instructed to complete the LOT-R and a measure of depression

at three time points (baseline, 1-year follow-up, and 4-5 year follow-up). The authors

found that LOT optimism baseline scores could predict depression at both follow-ups

in the first four younger-age groups. LOT-R pessimism predicted depression at the

two follow-ups in the first three younger-age groups.

The genetic and environmental origins of the links between dispositional

optimism and depression have been investigated in some studies. For instance, Plomin

et al. (1992) administered measures of dispositional optimism, depression, and life

satisfaction in 500 twins (72 pairs of identical twins reared apart, 126 pairs of

identical twins reared together, 178 pairs of fraternal twins reared apart, and 146 pairs

of fraternal twins reared together). It showed that both LOT optimism and LOT

pessimism were significantly associated with depression (r = -.31 and r = .44

respectively).



7.4 How to manipulate optimism?

Optimism-enhanced manipulations have been developed on the basis of the main

optimism approaches and implemented in numerous studies both in normal

populations and in clinical settings. Given the strong association between optimism

and depression, optimism interventions have been developed to promote optimistic

explanatory style and favourable expectations.

CBT-based optimism intervention: attributional retraining (AR)

In addition to Peterson et al. (1982)’s theory of attributional style, several other

attribution theories have been proposed. For example, the causal attribution theory of

Weiner (1985) specifically analyses the attributional style of students who are

vulnerable when searching for explanations of academic success and failure within

themselves, especially for negative events. According to Weiner’s proposal, all

attributions can be made along three dimensions: internality, stability, and

controllability. This 2 × 2 × 2 taxonomy offers eight possible causal attributions in

which any given explanation can be classified (Weiner, 1985).

Based on Weiner’s theory, attributional retraining (AR) has been developed to

help people to alter their maladaptive attributional style, reframe the way they think

about positive and negative life events, and develop more adaptive and self-helping

explanations for success and failure (Haynes, Perry, Stupnisky, & Daniels, 2009).

Most of the recent studies on AR have been conducted with college students, in whom

AR was found to have beneficial effects on cognition and academic performance (for

a review, see Haynes et al., 2009).



Self-administered optimism training (SOT)

In addition to mainstream AR manipulations, Fresco et al. (2009) developed self-

administered optimism training (SOT) based on traditional Cognitive Behavioural

Therapy (Beck, 1976), the reformulated learned helplessness theory (Abramson et al.,

1978), and the AR protocols, aiming to reduce current levels of pessimistic

explanatory style which are believed to predict depressive symptoms (Metalsky et al.,

1993). Theoretically, SOT represents an AR intervention that emphasizes a person’s

attention to daily life events and their explanations for these events by means of daily

writing (Fresco et al., 2009).

During a typical SOT session designed by Fresco et al. (2009), participants are

instructed to spend around 10 minutes each day for a week to identify 5 positive and 5

negative events in their life, finding initial causes along the dimensions of internality,

stability, and globality for each event, then revise and reassess alternatives and more

adaptive attributions for these events along the same three dimensions after reflection.

The process is completed within 28 days. The SOT was found to be effective in

building an optimistic explanatory style and reducing depressive symptoms in at least

some college students who scored high in attributions for negative events (Fresco et

al., 2009).

Other CBT-based optimism intervention techniques

In addition to AR techniques, a variety of other CBT-based optimism interventions

have been developed. For example, Burns (1980) proposed the anti-pessimism sheet

technique, which targets the specific expectations an individual holds for a relevant

situation.

Riskind and colleagues (1996) contributed several optimism interventions

designed specifically to decrease pessimism or increase optimism. One such



intervention helps the client to identify negative thinking and adopt a more adaptive

positive view. Positive visualization, which instructs the client to visually rehearse

attaining a positive outcome for a chosen negative event, was proposed as an

alternative technique for increasing optimism. The silver lining technique which was

described in this paper can be implemented more easily. Clients are instructed to

identify one genuinely positive element in one problematic situation. The technique of

pump priming was developed based on the principle of cognitive priming. This

technique aims to increase an individual’s ability to think and define situations

optimistically by priming the instantaneous approachability to working memory of

cognitive divisions that are demanded for optimism.

Positive writing and Best Possible Self （BPS）

King (2001) conducted a pioneering study in which participants were asked to

“imagine that everything has gone as well as it possibly could” (the Best Possible Self

condition, the BPS) and write about it for 20 minutes each day for four consecutive

days. This manipulation has been shown to be beneficial for promoting subject well-

being and has been replicated in two follow-up studies (Burton & King, 2004, 2008).

Within a group of third-year medical school students, the beneficial effects of writing

about emotions and goals were reported as well (Austenfeld et al., 2006).

Based on King’s study, the BPS imaginary exercise has been further used in

later studies of optimism intervention by many psychologists, with some alterations.

In the BPS intervention, participants normally are instructed to imagine and write

down some features (such as in the professional domain) that their future best possible

self should have. The interventions vary in time (from four days to four weeks), style

(writing or talking), administration (self-conducted or supervised by administrators),

and form (face-to-face or online).



Theoretically, the BPS manipulation aims to (temporarily) increase positive

expectations for the future by means of an experimental manipulation, which is

related to the beneficial effects of dispositional optimism (Meevissen et al., 2011).

The mechanism underlying the beneficial effects of BPS on well-being was assumed

to be the optimists’ tendency to generate more vivid positive mental images of future

events than pessimists (Blackwell et al., 2013). Evidence from the neurobiological

study of optimism partly supports this assumption. Brain images reveal that optimism

is associated with greater activation of a brain area that is related to positive imagery

of future events (Sharot, Riccardi, Raio, & Phelps, 2007).

Semantic optimism priming

Semantic optimism priming was used to temporarily manipulate generalized

expectations in one study conducted by Fosnaugh et al. (2009). Participants were

given a packet of scrambled sentence tests including 15 items (11 of which were

related to optimism), and told to build a sentence with four of the five words

contained in each item. It was assumed that this manipulation would activate

optimistic thinking unconsciously. It revealed that this optimism intervention is

effective in promoting dispositional optimism.

7.5 Empirical studies of optimism interventions

Optimism has long been seen as a simple yet powerful way for a person to cope more

adaptively with stress (Nes & Segerstrom, 2006; Scheier & Carver, 1992). Though

optimism interventions have been mainly integrated with other positive activities in

most previous practices, single optimism-enhanced manipulations have been

conducted both in non-clinical populations and in clinical settings. Generally, research

has shown that optimism interventions are effective in enhancing well-being and



reducing negative emotions (Austenfeld et al., 2006; Burton & King, 2004; Fosnaugh

et al., 2009; Littman-Ovadia & Nir, 2014; Meevissen et al., 2011).

7.5.1 Optimism interventions in nonclinical samples

Perry and colleagues have conducted a series of AR studies in college students

focusing on academic achievement (Haynes, Ruthig, Perry, Stupnisky, & Hall, 2006;

Perry, Hechter, Menec, & Weinberg, 1993; Perry & Penner, 1990; Ruthig, Perry, Hall,

& Hladkyj, 2004). In one of these studies (Ruthig et al., 2004), attribution retraining

was designed to improve academic motivation and achievement striving. The authors

found that the AR treatment group exhibited significantly lower test anxiety and

greater persistence in college courses than the control group. These types of studies

have shown that AR treatments are effective in fostering adaptive attibutional

thinking, positive academic motivation, and good academic performance (Haynes et

al., 2009). Riskind et al. (1996) introduced several AR-similar optimism training

methods and conducted these techniques in their study. They found that the optimism

training group reported more optimistic explanations, higher problem-solving self-

efficacy, and more positive cognition than the control group.

Because AR has been designed primarily to enhance student persistence

following possible academic failures, it has long been used to cultivate students’ more

adaptive attributions. The typical AR intervention instructs children to make a more

adaptive attribution, like lack of effort, instead of more pessimistic ones, like a lack of

ability, to their failure on academic tasks (Cecil & Medway, 1986). Most attribution

retraining techniques are more accessible to younger children compared with CBT-

based interventions, since they are much less cognitively demanding than cognitive

restructuring tasks. AR conducted in children has benefits in enhancing children’s

persistence in math problem-solving (Okolo, 1992), social competence (Aydin, 1988),

and reading tasks (Fowler & Peterson, 1981). However, the long-term effect of AR is



unknown; positive attributions may be hard to maintain if children are frequently

faced with failures.

In addition to AR, other optimism manipulations have also been applied in

normal populations. For example, The benefits of positive writing life goals was

compared with expressive talking about life goals in one study (Harrist, Carlozzi,

McGovern, & Harrist, 2007). Comparing with the control group, both intervention

groups reported less negative emotions, and writing intervention was more effective

in enhancing positive emotions. Sheldon and Lyubomirsky (2006) revealed that the

BPS intervention is more beneficial than the gratitude treatment for increasing and

maintaining positive emotions.

Peters and colleagues adapted the original BPS technique and conducted a

series of studies of BPS intervention. Their studies employed a random-assignment,

placebo-controlled design, in which participants in the optimism intervention

condition imagined and wrote about their future best possible self in a personal, a

relational, and a professional domain, for five-minute intervals per day over a period

of two weeks. Participants in the control group imagined and wrote down their daily

activities (Peters, Flink, Boersma, & Linton, 2010). In one study (Peters et al., 2010),

the BPS group exhibited larger increases in positive affect and positive future

expectations compared with the control group. BPS imagery caused a boost in

optimism, and the effects remained two weeks after the intervention ended. This result

was replicated in another study conducted by Meevissen & Peters (Meevissen et al.,

2011).

The benefits of thinking and writing optimistically were also replicated in

longer-term follow-up studies. For example, in one eight-month-long experimental

study, participants imagined and wrote their future BPS for 15 minutes a week over a

period of eight weeks. Individuals in the control condition listed what they did in the



previous seven days for 15 minutes a week. Notably, significant differences in

happiness between the intervention and comparison groups remained even six months

later (Lyubomirsky et al., 2011).

Evidence from BPS conducted online also supports its benefits in improving

psychological well-being. For example, Shapira and Mongrain (2010) conducted an

on-line intervention study, in which participants were randomly allocated into three

groups (BPS was one of the two intervention groups). The results showed that

individuals in the optimism condition were less depressed for up to three months and

were happier up to six months later compared to participants in the control condition.

Even self-administered optimism-cultivation activity is beneficial in reducing

negative emotions. For example, Littman-Ovadia and Nir (2014) adapted the three-

good-thing intervention to a brief daily self-administered optimism intervention,

which instructed the participants to “Think of three good things (items, people or

events) waiting for you tomorrow. Write them down. Choose one of them and try to

experience and maintain the sincere heart-felt feelings associated with it for five

minutes”. The intervention group did this practice for six consecutive days. This daily

optimism intervention effectively reduced pessimism, negative affect, and emotional

exhaustion at post-test and one month follow-ups.

In a study with undergraduate students, two different optimism manipulations,

optimistic orientation and optimism priming, were examined. It was found that both

interventions produced modest increases on a dispositional optimism measure and a

situational optimism measure, unlike in the control group (Fosnaugh et al., 2009).



7.5.2 Optimism intervention in clinical settings

Optimism intervention studies for alleviating depressive symptoms

Though diverse optimism interventions have been shown to be effective in promoting

positive emotions and reducing pessimism, very few optimism cultivation studies

have been conducted to directly decrease depressive symptoms. In the following three

rare examples, Self-Administered Optimism Training (SOT) and Attibutional

Retraining (AR), which have been developed based on the attributional theory of

depression, have demonstrated promising results in treating depression. In addition,

an adapted online optimism intervention study has also shown that positive optimism-

enhanced activities are effective in reducing depressive symptoms.

Fresco et al. (2009) randomly assigned 112 participants with a pessimistic

explanatory style and depressive symptoms (measured by BDI) into a SOT

experimental group or a no-treatment control group. Individuals in the intervention

group received 10 minutes of instruction concerning self-administering of optimistic

explanatory style, and then engaged in self-administered optimism training every day

for 28 days, while participants in the control condition were not involved in any tasks.

Participants in the intervention group reported a significant drop in their depressive

symptoms.

Sharifi, Hajiheidari, Khorvash, and Mirabdollahi (2013) examined the

effectiveness of a six-week attributional retraining intervention (two sessions per

week, forty-five minutes per session) on reducing depression and anxiety in 32

women who suffered from miscarriage. Participants were randomly assigned to either

an intervention group or a control group. Depression and anxiety were assessed at

three time points: pre-test, post-test, and five-week follow-up. Results demonstrated



that participants in the intervention group scored lower in depressive symptoms than

their counterparts in the control group both in the post-test and the follow-up.

Optimism interventions conducted online have also been shown to be

beneficial in ameliorating depressive symptoms. Sergeant and Mongrain (2014)

conducted an online optimism intervention over a period of three weeks and collected

two-month follow-up data. Participants (n = 466) were randomly assigned to the

optimism intervention group or the control group. Participants in the intervention

group were instructed to perform several optimism techniques, including “listing five

things that made them feel like their life was enjoyable, enriching, and/or worthwhile”,

listing “three things that could help them see the bright side of a difficult situation”,

and describing briefly a goal that “they would like to achieve in the next day or two”

with “steps they would like to meet this goal”. By contrast, participants in the control

condition were asked to describe their daily activities. Depression, dispositional

optimism, and happiness were measured. Results indicated that online optimism

cultivation practice was effective in decreasing depressive symptoms and promoting

happiness immediately and in the one- and two-month follow-ups, especially for

pessimists.

Optimism interventions in other clinical samples

Stanton et al. (2002) carried out a pioneering study on the written expression of

positive emotions within a group of breast cancer patients. The participants were

instructed to join a four-session writing task, including writing about their “positive

thoughts and feelings regarding their experience with breast cancer”. Patients who

wrote about the positive consequences of their experience had significantly fewer

negative physical symptoms and fewer medical appointments for cancer-related

morbidities at three months than did the control group. This finding was duplicated in



a later study conducted within another group of breast cancer patients (Low, Stanton,

& Danoff-Burg, 2006).

7.5.3 Optimism interventions in children and adolescents

Most studies on optimism intervention are conducted on adults, though there are still

some attempts in cultivating optimism and preventing depression in childhood and

adolescence. One such attempt, the Penn Resiliency Program (PRP), (Jaycox, Reivich,

Gillham, & Seligman, 1994) is comprised of cognitive-behavioral based interventions

targeting early adolescence (11-14 years old). Teachers and counselors at school

deliver this program. Intervention techniques have been adapted from adult CBT

(Beck, 1976), including self-disputing, goal setting, assertiveness, and negotiation

training. All these intervention techniques aim to help children to learn to challenge

their pessimistic explanatory style and develop adequate problem solving skills in

social life (Gillham & Reivich, 2004).

PRP has been shown to be effective in reducing moderate to severe depressive

symptoms after a two year follow-up (Gillham, Reivich, Jaycox, & Seligman, 1995).

Children who had completed the PRP were more inclined to show an optimistic

attributional style and less likely to be depressed compared with the control group

(Gillham & Reivich, 2004; Gillham et al., 1995). Results of several studies conducted

in Chinese samples also support the beneficial influence of the PRP in reducing

depressive symptoms and cultivating optimistic explanatory style in children (Yu &

Seligman, 2002). However, the effectiveness of the PRP has been challenged, since

some of the participants (one of the three schools) reported no significant decrease in

depressive symptoms after a three-year follow-up (Gillham et al., 2007). Cultivating

optimism techniques should be conducted with caution, considering the potential

influences of other individual and social factors.



Summary

Taken together, a diverse amount of optimism interventions have emerged to provide

possible answers to the question, how does one enhance well-being and relieve

suffering? AR and SOT were designed and developed based on attributional theories.

Generally, AR has been mainly conducted in academic backgrounds and has shown

beneficial effects on academic performance. By contrast, SOT aims to reduce current

levels of pessimistic attributional style that characterise depression. The BPS aims to

increase positive expectations which can be effective in boosting positive emotions

and in turn decreasing depressive symptoms.

7.6 Research questions

Using optimism interventions to decrease depressive symptoms

Traditionally, Cognitive Behavioural Therapy emphasised the influence of specific

beliefs and thoughts instead of focusing on broad cognitive biases such as explanatory

style and dispositional optimism, without examining the possibility of individual

differences in optimism (Pretzer & Walsh, 2001). The situation has recently changed

since psychologists began to understand optimism from a cognitive perspective, and

therefore including the approach of attributional style and dispositional optimism.

Previous research has shown that both SOT and BPS are effective in

promoting psychological well-being and reducing depressive symptoms. Applications

of these two optimism manipulations in empirical studies have yielded positive results

confirming the benefits of optimism interventions on enhancing well-being. However,

very little systematic work has been done to investigate the advantageous effects of

optimism interventions on psychotherapy applications in concrete settings. The



question of how to convert the benefits of optimism interventions to systematic and

effective activities diminishing depressive symptoms has not been addressed

adequately. Optimism intervention studies aimed particularly at decreasing depressive

symptoms and those clinically diagnosed with depressive disorders are needed.

Additionally, manipulating optimism has been conducted separately, aimed at

addressing general expectations or attributional style. There is no research including

both kinds of optimism interventions conducted so far to my knowledge. Since

previous research has shown that both optimism techniques are effective in promoting

psychological well-being and reducing depressive symptoms and theoretical

connections between attributional style and dispositional optimism have been found in

our early-stage analysis, the possibility of combining both SOT and BPS in one

optimism intervention study raises the possibility of fully understanding the

effectiveness of optimism interventions in depression treatment.

Participants: first-year college students

For my study of optimism interventions, young adults entering their first year of

university were chosen as targeted participants. Maladaptation of freshmen to

university life has been given much attention recently. Starting college is a

challenging time for first-year students and is often characterized by negative

emotions, such as depression and anxiety, which can negatively affect quality of life

and academic performance. First-year students typically experience a stressful life due

to a variety of causes, such as the challenges of living in a different and unfamiliar

environment (Negovan & Bagana, 2011). This life transition from late adolescence to

early adulthood may bring a series of difficult situations to deal with.

All these factors may increase first-year students’ vulnerability to depression.

Brandy, Penckofer, Solari-Twadell, and Velsor-Friedrich (2015) reported that 45% of



students demonstrated greater than average levels of stress and 48% reported

clinically significant depressive symptomology in one freshmen sample (N = 188). In

a sample of veterinary medical students (N = 240), data showed that 49%, 65%, and

69% of the participants reported depression levels at or above the clinical cut-off

across their first three semesters of study. Results indicated that transitional stress

predicted increased depression and anxiety symptoms and decreased life satisfaction

(Reisbig et al., 2012).

Some research has begun to investigate the role of optimism in psychological

adjustment during life transitions such as this. For example, Brissette et al. (2002)

reported that higher levels of dispositional optimism, assessed at the beginning of the

first semester of university, was prospectively associated with smaller increases in

stress and depression over the course of the first semester. Chemers et al. (2001)

found that LOT scores were strongly correlated with academic performance and

personal adjustment in a sample of first-year university students (N = 256). Similarly,

in a much larger sample of college freshmen (n = 2,189), L. S. Nes et al. (2009) found

that optimistic students had better psychological adjustment and motivation than

pessimists in the period of college transition. Students with a higher level of

dispositional optimism were more likely to return to school for the second year, with

increased motivation and decreased distress.

Though there is no single study that has directly examined the relationship

between attributional style and depression in first-year college students, it has been

reported that students who had pessimistic attributions for their academic failure

received lower exam scores than their freshmen counterparts who held an optimistic

attributional style in explaining academic failure (Peterson & Barrett, 1987).

Academic stress has been found to be a strong predictor of depression and anxiety in a

group of veterinary medical students during their first three semesters (Reisbig et al.,



2012). Based on these findings concerning the influences of attributional style and

dispositional optimism on academic performance, depression, and psychological

adjustment in first-year college students, interventions targeting cultivating optimism

in this specific group should be considered for decreasing depressive symptoms to

enhance their college experience.

My aim was to test whether manipulations based on optimism theories might

alleviate depressive symptoms in first-year college students. I hypothesised that

optimism interventions can produce stronger and lasting benefits on psychological

well-being, especially in reducing depressive symptoms of participants in the

experimental condition than in the control condition.


Chapter 8: Optimism interventions for depression in first-year college students 211

Chapter 8: Optimism interventions for depression in first-year college students

8.1 Study 1: individual optimism interventions with depression

8.1.1 Intervention designs

Corresponding respectively to dispositional optimism and explanatory style, two

optimism manipulation techniques were adopted in my interventions for depression.

One is the Best Possible Self (BPS) technique adapted from several previous BPS

studies. As in the BPS intervention, participants normally are instructed to imagine

and write down some aspects (such as professional domain) that their future best

possible self should have. The interventions were variant in time (from 4 days to 4

weeks), style (writing or talking), administration (self-conducted or supervised by

administrators), domains of writing (three or more), and form of intervention (face to

face or online). Borrowing from Lyubomirsky et al. (2011)’s BPS paradigm, students

in my study were instructed to write about their best possible future in each of the 7

domains (romantic life, educational attainment, hobbies or personal interest, family

life, career situation, social life, and physical/mental health). Instead of doing BPS

every week, students were asked to do these positive writings on a daily basis across

a week, similarly to the BPS study of Peters, Meevissen, and Hanssen (2013).

The other optimism activity is the self-administered optimism training (SOT)

adapted from Fresco et al. (2009). In their SOT study, participants were instructed to

spend around 10 minutes each day for a week to identify five positive and five

negative events in their life, finding initial causes along the dimensions of internality,

stability, and globality for each event, then revising and reassessing alternative and

more adaptive attributions for these events along the same three dimensions after

some reflection. The whole procession of SOT is completed within 28 days. We

adapted Fresco et al. (2009)’s SOT into a shorter version of 7 days. Instead of

identifying five positive and five negative events in their life each day, participants

are asked to identify three positive and three negative events.



Previously, SOT was applied in one study aiming at reducing depressive

symptoms, and BPS was adopted only in intervention studies of nonclinical samples

to my knowledge. In addition to SOT and BPS, face-to-face individual

psychotherapy was conducted in my study on the basis of individual positive

psychotherapy with mild-to-moderate depression (Seligman et al., 2006). In total, the

whole intervention consisted of three sessions, in which each consists of a 45-

minutes face-to-face individual counselling.

The first session is SOT practice. Before SOT, every participant in the

intervention group receives an individual counselling, in which the counsellor

introduces basic theory of attributional style, and gives instructions of the SOT

procedure. Then the participant is asked to do homework. The homework contains

approximately 15 minutes of SOT every day in the following week. The daily SOT

is completed following three steps: (a) self-monitoring daily 3 negative and 3

positive events; (b) identifying the initial cause, and rating that cause along the

dimensions of internality, stability and globality; (c) brainstorming additional or

alternate causes; and (d) arriving at a revised cause that was also rated along the

dimensions of internality, stability, and globality.

The second session is BPS exercise. Similar as SOT session, every participant

in the intervention group receives a 45-minute individual counselling, in which the

counsellor helps the participant identify their core values, they were asked to “think

about how they wanted to be remembered at the end of their lives by their loved ones”

(Peters et al., 2013). Home work is assigned at the end of the individual counselling.

The participant is asked to imagine and write about his or her life if everything

unfolded as he or she wanted. The participant is instructed to envisage that perhaps

he or she has worked diligently and achieved his or her most important dreams. Once

this image had been invoked, the participant wrote about this future for 15 minutes in

one of 7 aspects, including best possible future romantic life, educational attainment,

hobbies or personal interest, family life, career situation, social life, and

physical/mental health. These tasks were required to be completed on a daily basis in

the following week.



The third session is for summary and post-intervention test. The counsellor

and the participant review progress of intervention and discuss gains and

maintenance of these two positive activities. At the end, the participant completes

measures of depression, attributional style, dispositional optimism, and subjective

well-being (life satisfaction).

Hypotheses

Our first hypothesis concerned the beneficial effects of optimism intervention on

depressive symptoms. I predicted that participants in the experimental group would

experience lower levels of depression outcomes by the end of the intervention than

the control group, and that these beneficial effects might even be maintained at the

one-month and three-month follow-ups.

Similarly, our second hypothesis was that for the intervention group, a

decrease in depressive symptoms would be accompanied by the corresponding

improvement in optimistic explanatory style, especially for attributions of negative

events, not only immediately after the manipulations, but also the following three

months after the interventions had ended.

Also, I predicted that positive activities would bolster subjective well-being

(life satisfaction) and dispositional optimism and decrease dispositional pessimism

immediately after the intervention, and these improvements might last in the follow-

up periods.

8.1.2 Method

Participants

Fifty-two undergraduate students in Sample 5 (see Chapter 1.5.4 for details) took

part in this study. All participants were native Chinese speakers with ages ranging

from 17 to 21 (M = 18.50, SD = 0.71). They were randomly divided into one of the

two conditions: an experimental group (n = 26) and a control group (n = 26). There

were no significant differences in gender, age, ethnicity, and year of education

between these two conditions.



Not all participants completed the whole procedure. Three participants

dropped out of the intervention group and two dropped out of the control group. As a

result, there were 23 participants in the intervention group and 24 participants in the

control group available for the final data analysis (M = 19.07, SD = 0.86; 19 males

and 28 females). There were no significant differences in gender, age, ethnicity,

years of education, or pre-test measures between those who remained in this study

and those who left.

Measures

Attributional style was measured using a Chinese version of the ASQ (Zhang, 2006).

The ASQ takes on average 15 minutes to complete. Composite attributional styles

were calculated separately for positive and negative events. Higher scores for

positive events and a lower score for negative events on any area demonstrates a

more “optimistic” attributional style for that domain, i.e., more external, temporary

and specific for negative events, and more internal, stable and global for positive

events. Cronbach’sαof the pre-test for the scale was 0.85 for negative events and

0.66 for positive events; for the post-test, 0.82 for negative events and 0.89 for

positive events; and for the three-mohth follow-up, 0.86 for negative events and 0.86



Orientation Test-Revised (Lai & Yue, 2000). Subjects were scored for two separate

composite scores, LOT-R Optimism and LOT-R Pessimism. Cronbach’sαfor the

pre-test was 0.74 for dispositional optimism and 0.62 for dispositional pessimism; for

the post-test, 0.47 for LOT-R Optimism and 0.72 for LOT-R Pessimism; for the one-

month follow-up, 0.74 for LOT-R Optimism and 0.62 for LOT-R Pessimism; and for

the three-month follow-up, 0.62 for LOT-R Optimism and 0.52 for LOT-R

Pessimism.

Subjective well-being was assessed using a Chinese version of Satisfaction

with Life Scale (SWLS; Chen & Zhang, 2004). Subjects were scored for total



optimism scores. Cronbach’sαfor the pre-test was 0.82; for the post-test, 0.83; for

the one-month follow-up, 0.85; and for the three-month follow-up, 0.80.

A Chinese version of the Beck Depression Inventory (BDI; Chan & Tsoi,

1984) was used to measure depression. Cronbach’sαfor the pre-test was 0.83; for

the post-test, 0.82; for the one-month follow-up, 0.87; and for the three-month

follow-up, 0.83.

Procedure

Recruiting participants. To conduct the present optimism intervention pilot

study, a general sample was recruited from all 980 freshmen in China Youth

University of Political Studies. To test mental health of first-year students, Self-

Reporting Inventory 90 (SCL-90; Derogatis & Cleary, 1977; Derogatis, S, Covi, &

Rickeis, 1973) was conducted in the end of the first month of their entry into the

university. According to the generally accepted criterion of SCL-90, a total score of

160 and above or a score of 2 and above for any single dimension was seen as

indicators of possible mental illness. Accordingly, a total SCL-90 score of 160 or

above and a score of 2 or above in depression were utilized in selecting eligible

participants. A total of 85 students were selected as a general sample based on the

criterion above and were contacted by teachers of the University Consulting Centre.

The research was presented as a study involving activities designed to develop

personal strength and psychological well-being. Finally, 52 students agreed to take

part in this study.

Baseline assessment. Participants completed the first set of questionnaires at

their convenience within a week. Baseline assessments included a consent form,

demographic questions, and measures of depression, attributional style, dispositional

optimism, and SWB (life satisfaction). The consent form informed students of their

rights as participants in this study. They then were asked to provide general

background information, such as gender, ethnicity, age, and married status. Three

days after completion of the baseline questionnaires, participants began the

intervention.



Optimism interventions. Students were randomly assigned to either an

experimental condition or a control condition for a period of up to three weeks.

For the experimental condition, optimism interventions took place over three

sessions (each session lasts for about 40 minutes) over three consecutive weeks.

Individual face-to-face counselling was conducted by three qualified counsellors in

the University Consulting Centre. They followed the intervention manual to conduct

all the intervention sessions. A notebook was assigned to participants in the

intervention group in the first session for completing their homework. The

homework can be written down on the notebook or be printed out. In the intervention

period, participants in the control group were not involved in any tasks related to this

study.

Time 1, time 2, and time 3 assessments. Optimism intervention participants

completed the measure battery in the final session, and control participants were

scheduled a similar time for their Time 1 measure. Then participants in both

conditions were scheduled a time to return for their Time 2 (one-month follow-up),

and Time 3 (three-month follow-up) packet of self-report measures. Because of the

length of time it took to take the questionnaire (approximately 15 minutes), the ASQ

was only re-administered at Time 1 and Time 3.

8.1.3 Results

Baseline descriptive

An independent samples t-test on baseline scores between the intervention group and

control group revealed no significant differences between the two groups on any of

the measures (LOT-R Optimism, LOT-R Pessimism, ASQ Negative, ASQ Positive,

SWLS, and BDI), indicating that randomization was successful.

Table 8.1 shows the descriptives and correlations of baseline scores for the

whole sample on the LOT-R, ASQ, SWLS, and BDI. In line with at least one

previous finding (Isaacowitz & Seligman, 2002), both ASQ Negative and ASQ

Positive did not significantly correlate either LOT-R Optimism or LOT-R Pessimism,

indicating that explanatory style of life events may be uncorrelated to general



expectancies of future events. As expected, the LOT-R Pessimism was positively

correlated with BDI (r = 0.29) and negatively correlated with SWLS (r = -0.41), and

BDI was negatively correlated with SWLS (r = -0.38).

Measures Descriptives Correlations

Mean SD 1 2 3 4 6

1. BDI 20.60 8.73

2. LOT-R Optimism 6.28 2.50 -0.25

3. LOT-R Pessimism 5.20 2.22 0.29* -0.41**

4. SWLS 15.66 6.10 -0.38** 0.13 -0.19

5. ASQ Negative 13.64 2.06 -0.07 0.06 0.24 0.09

6. ASQ Positive 15.15 1.42 0.28 -0.05 0.24 -0.17 0.07

Table 8.1: Descriptives and intercorrelations between measures at baseline.

* p < 0.05. ** p < 0.01.

Intervention effects: immediate and longer term changes

Means and standards deviations for all measures for both two conditions from

baseline to post-interventions, as well as to one-month follow-up and three-month

follow-up are presented in Table 8.2. Changes for all measures for both groups in

four time-points are illustrated in Figures 8.1-8.6 (based on standardized scores).



Measures Pre-test Post-test 1-month Follow-up 3-month Follow-up

Mean SD Mean SD Mean SD Mean SD

Intervention group

BDI 20.09 8.90 11.39 6.82 11.57 6.85 11.00 4.94

LOT-R Optimism 6.26 2.61 8.09 1.78 7.91 2.25 7.57 1.16

LOT-R Pessimism 5.17 2.25 4.48 1.75 4.43 1.85 4.65 1.56

SWLS 15.61 6.29 19.70 4.95 21.39 5.68 19.35 6.03

ASQ Negative 13.59 1.86 12.99 1.88 12.64 2.20

ASQ Positive 15.31 1.44 16.11 1.45 15.49 1.69

Control group

BDI 21.08 8.72 21.58 5.56 17.33 8.05 18.04 8.30

LOT-R Optimism 6.29 2.44 7.29 2.05 6.92 2.06 7.08 1.79

LOT-R Pessimism 5.42 2.22 5.33 2.24 5.25 1.98 4.92 1.86

SWLS 15.71 6.05 18.33 6.65 19.67 6.95 18.42 4.66

ASQ Negative 13.69 2.28 13.82 1.75 14.22 1.80

ASQ Positive 14.99 1.42 14.64 2.09 14.85 1.94

Table 8.2: Means and Standard Deviations of outcome measures by condition at all

time-points.

Figure 8.1: Depression as measure by the BDI at baseline, at post-intervention, 1-

month follow-up, and 3-month follow-up per condition.

-0.80

-0.60

-0.40

-0.20

0.00

0.20

0.40

0.60

0.80

Pre Post 1m 3m

intervention group

control group



Figure 8.2: Dispositional Optimism as measure by the LOT-R at baseline, at post-

intervention, 1-month follow-up, and 3-month follow-up per condition.

Figure 8.3: Dispositional Pessimism as measure by the LOT-R at baseline, at post-


-0.50

-0.40

-0.30

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

0.50

Pre Post 1m 3m

intervention group

control group

-0.30

-0.20

-0.10

0.00

0.10

0.20

0.30

Pre Post 1m 3m

intervention group

control group



Figure 8.4. Subjective well-being as measure by the SWLS at baseline, at post-


Figure 8.5. Attributional style for negative events as measure by the ASQ at baseline,

at post-intervention and 3-month follow-up per condition.

-0.60

-0.40

-0.20

0.00

0.20

0.40

0.60

Pre Post 1m 3m

intervention group

control group

-0.50

-0.40

-0.30

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

Pre Post 3mIntervention Group

Control Group



Figure 8.6: Attributional style for positive events as measure by the ASQ at baseline,


Immediate post-intervention

Right after the completion of the three-week intervention, supporting our first

hypothesis, students in the experimental group reported a greater decrease in

depressive symptoms relative to students in the control group (see Figure 8.1), t(45)

= -5.63, p < .001. However, although participants in the intervention group displayed

a tread toward a greater increase in LOT-R Optimism and a greater decrease in LOT-

R Pessimism relative to the control group right after the intervention (see Figure 8.2

and Figure 8.3), two-tailed t tests showed that the experimental group and the control

group did not significantly differ on either LOT-R Optimism or LOT-R Pessimism.

Similarly, as displayed in Figure 8.4, although intervention group participants were

still showing a trend toward greater subjective well-being gains compared with the

control group, it was not significant.

Participants in the experimental group reported a greater increase in

explanatory style for positive events relative to participants in the control group (see

Figure 8.6), t(45) = 2.79, p = .008. However, comparison of the ASQ-negative

-0.40

-0.30

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

0.50

0.60

Pre Post 3m

Intervention Group

Control Group



contrasting the intervention group with the control group failed to reach statistical

significance, though participants in the experimental group displayed a trend toward

a decrease in explanatory style for negative events while the control group displayed

a trend toward increase (see Figure 8.5).

Follow-ups

As expected, again supporting our most important prediction, depression scores of

the intervention group were much lower than those in the control group, t(45) = -2.64,

p = .01, though depressive symptoms in the control group also experienced a trend of

slight decrease (see Figure 8.1); and this significant difference was even bigger three

months after the intervention had ended, t(45) = -3.52, p = .001.

The comparison of LOT-R Optimism, LOT-R Pessimism, and SWLS

contrasting the experimental group with the control group failed to reach statistical

significance in either the one-month follow-up or the three-month follow-up.

Although participants who had completed the optimism intervention displayed a

trend toward greater increases in life satisfaction relative to the control group, one

month after the intervention had ended (see Figure 8.4), this difference did not reach

statistical significance. For LOT-R Optimism, students in the intervention group

showed trend of decreas one month and also three months after the intervention had

ended, while their counterparts in the control group were showing a trend of losses in

the one-month follow-up and then a trend of gains in the three-month follow-up.

That is, the optimism scores of the control group in the one-month follow-up was

lower than in post-intervention, then the level of optimism increased in the three-

month follow-up compared with the one-month follow up. (see Figure 8.2). however,

scores of LOT-R Pessimism in the one-month follow-up were lower than in post-

intervention for both the experimental group and the control group. For the three-

month follow-up, the intervention group showed an increase in LOT-R Pessimism

scores, while the control group showed an increase (see Figure 8.3). The changes and

differences in both LOT-R Optimism and LOT-R Pessimism were not significant.



Explanatory style as measured by the ASQ showed different changing

patterns for positive and negative events three months after the optimism intervention.

Specifically, for the ASQ-Negative, students in the experimental group showed

decreased scores, while the control group students showed an increase (see Figure

8.5), and as a result the intervention group participants reported more optimistic

explanatory styles for negative events than their counterparts in the control group,

t(45) = -2.68, p = .01. However, a comparison between the intervention group and

the control group on the ASQ-positive failed to reach statistical significance. It

showed that participants in the experimental group displayed a trend toward

decreased scores, while the control group displayed a trend of slightly increased

scores (see Figure 8.6). Different changing patterns between explanatory style for

positive and negative events were consistent with previous findings of the ASQ

structure; attributional biases to positive events and to negative events emerged as

uncorrelated in the joint model (see Chapter 2.1).

8.1.4 Discussion

Results provided partial confirmatory support for the hypotheses. They indicate that

at post-intervention, one month and three months following the intervention,

individuals in the optimism condition were less depressed than those in the non-

treatment control condition. This provides preliminary evidence of the effectiveness

of optimism manipulations on reducing depressive symptoms. Data analysis also

revealed that positive activities in optimism were beneficial in developing optimistic

explanatory styles, especially for attributions for negative events. Overall, these

results are in line with previous findings that optimistic thinking can have

advantageous psychological benefits (Fresco et al., 2009; King, 2001). The results

indicate that these positive activities can lead to sustained increase in optimism and

decrease in depressive symptoms. Moreover, the effects remained one month and

three months later after the intervention had ended.

A number of potential active elements in the positive, future-oriented optimism

intervention may have contributed to these positive outcomes, such as the feeling of



attentive communication, positive re-evaluation of life events, and active arousal of

expectations.

Participants in the intervention group did not experience higher levels of

dispositional optimism or life satisfaction following the intervention period. There

are several possible reasons for this.

First, the sample size was rather small in total (N = 47). The results showed that

participants who had completed the optimism intervention generally displayed a

trend toward greater increases in dispositional optimism and life satisfaction relative

to the control group immediately and one month after the intervention, but the

differences did not reach statistical significance. Second, the general sample was

selected based on a total SCL-90 score of 160 or above and a score of 2 or above in

depression. Since depression and psychological dysfunction were utilized in

selecting eligible participants, it is possible that the optimism interventions may be

more effective for decreasing depressive symptoms than for increasing positive

feelings and general expectations, though benefits in decreasing depression have

been gained though boosting positive affections. Finally, it has been theoretically and

empirically widely accepted in positive psychology that relieving negative feelings

and increasing positive feelings are two separate endeavours (Seligman &

Csikszentmihalyi, 2000; Seligman et al., 2005).

Altogether, the current investigation indicated that optimism manipulations over

a period of two weeks led to significantly larger improvements in depressive

symptoms and increase in optimistic explanatory style compared to not receiving any

treatment. A different pattern emerged for short-term and long-term effects, such that

a relatively large reduction in depressive symptoms occurred immediately after the

intervention period, whereas the one-month and three-month follow-ups featured

stable levels of depression.



Further questions

In this pilot study, instead of conducting group optimism interventions that have been

mainly applied in previous research, an individual approach with face-to-face

counselling sessions was used. As results have shown, this individual intervention

was effective in decreasing depressive symptoms and in enhancing optimistic

attributional style. However, the way in which self-administered positive activities

and individual counselling were combined made it unclear what might be the cause

of those benefits. Whether it was the self-administered optimism interventions or the

individual consulting session is an unresolved question.

Additionally, the possibility of social desirability and demand effects when

students were keen on making good impressions to the counsellors might also be a

factor that should be considered. Moreover, though no-treatment control design has

been used in previous studies, it is more plausible to apply ‘placebo’-treatment

control design in intervention studies.

8.2 Study 2: group optimism interventions with depression

8.2.1 Intervention designs

Considering the unresolved questions from Study 1, I conducted a second study in

which purely self-administered optimism interventions were applied in first-year

college students with mild-to-moderate depressive symptoms. Two changes were

made in Study 2. The first was that the individual counselling sessions were excluded

in the experimental condition. The second change was that participants in the control

condition were asked to list their daily activities instead of doing nothing.

As in Study 1, optimism interventions in Study 2 consisted of two optimism

manipulation techniques, namely BPS and SOT. Participants were instructed to

complete SOT in the first week, and then complete BPS in the second week on a self-

administered basis.



Hypothesis

My first hypothesis concerned the beneficial effects of the optimism intervention on

depressive symptoms. I predicted that participants in the experimental group would

have lower levels of depression outcomes by the end of the intervention, and that

these beneficial effects might even be maintained at the one-month and three-month

follow-ups.

Similarly, my second hypothesis was that for the intervention group, a

decrease in depressive symptoms would be accompanied by a corresponding

improvement in optimistic explanatory style, especially for attributions of negative

events, not only immediately after the manipulations, but also following three

months after the interventions had ended.

Also, I predicted that our positive activities would bolster SWB (life

satisfaction) and dispositional optimism immediately and decrease dispositional

pessimism after the intervention and these improvements might last in the follow-up

periods.

8.2.2 Method

Participants

Participants in Sample 6 were involved in this study (see Chapter 1.5.4 for details).

Measures

Attributional style was measured using a Chinese version of the ASQ (Zhang, 2006).

Two composite scores, ASQ Negative and ASQ Positive, were calculated to assess

attributional style for negative and positive events respectively. Cronbach’sαfor the

pre-test for the scale were 0.73 for negative events and 0.84 for positive events; for

the post-test, 0.86 for negative events and 0.88 for positive events; and for the three-

month follow-up, 0.72 for negative events and 0.83 for positive events.


Orientation Test-Revised (Lai & Yue, 2000). Subjects were scored for two separate



composite scores, LOT-R Optimism and LOT-R Pessimism. Cronbach’sαfor the

pre-test for the scale was 0.50 for LOT-R Optimism and 0.53 for LOT-R Pessimism;

for the post-test, 0.47 for LOT-R Optimism and 0.59 for LOT-R Pessimism; for the

one-month follow-up, 0.61 for LOT-R Optimism and 0.64 for LOT-R Pessimism;

and for the three-month follow-up, 0.43 for LOT-R Optimism and 0.56 for LOT-R

Pessimism.

Subjective well-being was assessed using a Chinese version of the

Satisfaction with Life Scale (SWLS; Chen & Zhang, 2004). Subjects were scored for

total optimism scores. Cronbach’sαfor the pre-test for the scale was 0.79; for the

post-test, 0.74; for the one-month follow-up, 0.76; and for the three-month follow-up,

0.80.

Depression was measured using a Chinese version of the Beck Depression

Inventory (BDI; Chan & Tsoi, 1984). Cronbach’sαfor the pre-test for the scale was

0.79; for the post-test, 0.75; for the one-month follow-up, 0.70; and for the three-

month follow-up, 0.49.

Procedure

Participant recruiting and baseline assessment were the same as in Study 1.

Optimism interventions. Students were randomly assigned to either an

experimental condition or a control condition for a period of 2 weeks.

For the experimental condition, participants were instructed to apply SOT in

the first week, and then apply BPS in the second week. Participants reported to small

group (5-6 people per group) training sessions, which consisted of approximately 10

minutes of instructions on how to apply SOT and BPS in the beginning of the first

week and the second week. Participants were asked to complete their homework on a

self-administered basis (the same as in Study 1).



For the comparison control condition, participants were asked to spend 15

minutes per day listing what they did during that day. A notebook was assigned to

students for writing down their daily activities.

Time 1, time 2, and time 3 assessments. Optimism intervention participants

completed the measure battery in the following three days after they completed the

intervention sessions, and control participants were scheduled a similar time for their

Time 1 measure. Then participants in both conditions were scheduled a time to return

for Time 2 (one-month follow-up), and Time 3 (three-month follow-up) packet of

self-report measures. The ASQ was only re-administered at Time 1 and Time 3 due

to its length.

8.2.3 Results and analysis

An independent samples t-test on baseline scores between intervention group and

control group revealed no significant differences between the groups on any of the

measures (LOT-R Optimism, LOT-R Pessimism, ASQ Negative, ASQ Positive,

SWLS, and BDI). The descriptives and correlations of baseline scores for the whole

sample on the LOT-R, ASQ, SWLS, and BDI are shown in Table 8.3.

As shown in Table 8.3, BDI was negatively correlated with SWLS (r = -.35);

LOT-R Optimism was negatively correlated with LOT-R Pessimism (r = -.33) and

positively correlated with ASQ Positive (r = .30); and ASQ Negative was negatively

correlated with SWLS (r = -.31).



Measures Descriptives Correlations

Mean SD 1 2 3 4 5

1. BDI 21.81 7.40 -

2. LOT-R Optimism 8.19 1.76 -0.13 -

3. LOT-R Pessimism 4.41 1.58 0.23 - 0.33* -

4. ASQ Negative 12.74 1.60 0.07 -0.14 -0.04 -

5. ASQ Positive 14.07 1.90 -0.23 0.30* -0.20 -0.03 -

6. SWLS 16.22 5.76 - 0.35** 0.12 -0.21 - 0.31* 0.13

Table 8.3: Descriptives and intercorrelations between measures at baseline.

* p < 0.05. ** p < 0.01.

Intervention effects: immediate and longer term changes

Means and standards deviations for all measures for both conditions from baseline to

post-interventions, as well as to one-month follow-up and three-month follow-up are

presented in Table 8.4.

Changes for all measures for both groups in four time-points are illustrated in

Figures 8.7-8.12 (based on standardized scores).



Measures Pre-test Post-test 1-month Follow-up 3-month Follow-up

Mean SD Mean SD Mean SD Mean SD

Intervention group

BDI 21.43 7.44 17.47 6.03 16.27 5.04 16.63 3.80

LOT-R Optimism 8.17 1.90 9.13 1.04 9.17 1.53 8.80 1.58

LOT-R Pessimism 4.27 1.53 3.87 2.16 3.90 2.32 4.10 1.92

SWLS 16.13 5.89 19.53 4.47 20.20 4.78 19.60 3.66

ASQ Negative 12.59 1.72 11.83 1.88 - - 11.68 1.51

ASQ Positive 14.13 1.73 14.83 1.38 - - 15.04 1.88

Control group

BDI 22.21 7.46 21.10 6.12 19.48 5.16 18.90 4.93

LOT-R Optimism 8.21 1.63 8.24 2.28 8.34 1.42 8.31 1.65

LOT-R Pessimism 4.55 1.64 4.69 1.93 4.62 1.52 4.59 1.68

SWLS 16.31 5.73 17.59 4.08 18.24 5.14 19.07 5.59

ASQ Negative 12.90 1.49 12.95 2.13 - - 12.76 1.66

ASQ Positive 14.01 2.08 14.40 2.98 - - 14.31 1.96

Table 8.4: Means and Standard Deviations of outcome measures by condition at all

time-points.

Figure 8.7: Depression as measure by the BDI at baseline, at post-intervention, 1-

month follow-up, and 3-month follow-up per condition.

-0.60

-0.40

-0.20

0.00

0.20

0.40

0.60

Pre Post 1m 3m

intervention group

control group



Figure 8.8: Dispositional Optimism as measure by the LOT-R at baseline, at post-


Figure 8.9: Dispositional Pessimism as measure by the LOT-R at baseline, at post-


-0.30

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

Pre Post 1m 3m

intervention group

control group

-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

0.05

0.10

0.15

0.20

0.25

Pre Post 1m 3m

intervention group

control group



Figure 8.10: Subjective well-being as measure by the SWLS at baseline, at post-


Figure 8.11: Attributional style for negative events as measure by the ASQ at

baseline, at post-intervention and 3-month follow-up per condition.

-0.50

-0.40

-0.30

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

0.50

Pre Post 1m 3m

intervention group

control group

-0.50

-0.40

-0.30

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

Pre Post 3mintervention group

control group



Figure 8.12: Attributional style for positive events as measure by the ASQ at baseline,


Immediate post-intervention

Right after the completion of the two-week intervention, supporting the first

hypothesis, students in the experimental group reported a greater decrease in

depressive symptoms relative to students in the control group (see Figure 8.7), t(57)

= -2.30, p = 0.025. However, although participants in the intervention group

displayed a trend toward an increase in LOT-R Optimism and a decrease in LOT-R

Pessimism relative to the control group right after the intervention (see Figure 8.8

and Figure 8.9), two-tailed t tests showed that the experimental group and the control

group did not significantly differ on either LOT-R Optimism (t(57) = 1.95, p = 0.057)

or LOT-R Pessimism (t(57) = -1.54, p = 0.129). Similarly, as displayed in Figure

8.10, although intervention group participants were still showing a trend toward

greater subjective well-being gains compared with the control group, it was not

significant (t(57) = 1.75, p = 0.086).

For explanatory style measured by the ASQ, participants in the experimental

group reported a greater decrease in ASQ-Negative relative to participants in the

-0.30

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

Pre Post 3m

intervention group

control group



control group (see Figure 8.11), t(57) = -2.16, p = 0.035. However, a comparison of

the intervention group with the control group on the ASQ-Positive failed to reach

statistical significance, though participants in the experimental group displayed a

trend toward an increase in ASQ-Positive (see Figure 8.12).

Follow-ups

As expected, again supporting our most important prediction, BDI scores of the

intervention group were lower than those in the control group, t(57) = -2.42, p =

0.019, though depressive symptoms in the control group also experienced a trend

toward decreasing one month after the intervention (see Figure 8.7). This difference

was kept three months after the intervention had ended but did not reach statistical

significance, t(57) = -1.98, p = .053.

A comparison of LOT-R Optimism, LOT-R Pessimism, and SWLS

contrasting the experimental group with the control group in the one-month follow-

up or the three-month follow-up failed to reach statistical significance, with one

exception. LOT-R Optimism scores for the intervention group were significantly

lower than those in the control group one month after the intervention had ended,

t(57) = 2.13, p = 0.037, though LOT-R Optimism in the control group also

experienced a trend toward increasing (see Figure 8.8). For life satisfaction, although

participants who had completed the optimism intervention displayed a trend toward

greater increases relative to the control group one month after the intervention had

ended, this difference did not reach statistical significance (see Figure 8.10).

As expected, three months after the intervention had ended, ASQ-Positive

and ASQ-Negative showed beneficial changing patterns, though only the differences

and changes of ASQ-Negative reached statistical significance. Specifically, for the

ASQ-Negative, participants in the experimental group decreased their scores while

the control group kept a relatively stable level (see Figure 8.11), and as a result the

intervention group participants reported lower ASQ-Negative scores than their

counterparts in the control group, t(57) = -2.62, p = .011. For ASQ-Positive,

participants in the experimental group displayed a trend toward increasing their



scores, while the control group slightly decreased their scores. The difference of

ASQ-Positive scores between the intervention group and the control group in the

three-month follow-up was even bigger than the difference between these two groups

in post-intervention (see Figure 8.12). However, this difference failed to reach

statistical significance.

8.2.4 Discussion

The current investigation demonstrated that minimally supervised and self-

administered optimism interventions for a two-week period could result in decreases

in depressive symptoms and pessimistic explanatory style and enhance dispositional

optimism. Although participants in the experimental group did not significantly

decrease dispositional pessimism and significantly increase subjective well-being,

findings indicate that increases in dispositional optimism and decreases in

pessimistic explanatory style were associated with decreases in depressive symptoms.

Moreover, the benefits in decreasing depression in the intervention group

continued one month and three months after the intervention. These results indicate

that a brief and self-monitored intervention is effective in reducing symptoms of

depression and enhancing well-being.

8.3 General discussion

Both studies shared similar and slightly different trends in changes of depressive

symptoms in LOT-R Optimism, LOT-R Pessimism, ASQ Positive, ASQ Negative,

and subjective well-being of experiment groups. They generally showed a greater

increase in LOT-R Optimism, ASQ Positive, and subjective well-being and a greater

decrease in depressive symptoms, LOT-R Pessimism, and ASQ Negative for

participants in the intervention group than their counterparts in the control group,

though not all of these changes and differences reached statistical significance. For

example, though LOT-R Optimism showed a greater increase in post-intervention for

the intervention group in both studies, it produced differential increases between the

intervention condition and the control condition only in the one-month follow-up in

Study 2. This finding was unexpected given that previous findings showed that



writing about and imagining a BPS leads to an immediate increase in dispositional

optimism (Meevissen et al., 2011; Peters et al., 2010; Peters et al., 2013). Given that

previous studies of BPS had only been applied in non-clinical settings and

participants in my studies were first-year college students with mild-to-moderate

depressive symptoms, the failure of significant increases in LOT-R Optimism might

not be so unexpected.

I should point out, though, that the level of depression reduction of the

intervention group in Study 2 was lower as compared to the level of depression

reduction for the intervention group in Study 1. Two considerations may be helpful

to account for the smaller difference found between the experiment group and the

control group in Study 1 than in Study 2. First, it has been argued that individual

positive psychotherapy is effective in reducing depressive symptoms (Seligman et al.,

2006). Accordingly, since three individual counselling sessions were included in

Study 1 in addition to SOT and BPS exercises, and these individual counselling

sessions were excluded in Study 2, differences in reduction of depressive symptoms

between these two studies could be expected. Secondly, as noted by some

researchers, one of the major concerns in psychological assessment is the possibility

of social desirability and demand effects. The social desirability bias might be larger

if students were keen on making good impressions to the counsellors. Hence, it is

possible that participants in the intervention group in Study 1 were more obviously

affected by social desirability than in Study 2.

.


Chapter 9: Understanding optimism 237

Chapter 9: Understanding optimism

Thinking rosy futures is as biological as sexual fantasy. Optimistically calculating

the odds is as basic a human action seeking food when hungry or craving fresh air in

a dump. Making deals with uncertainty marks us [as a species] as plainly as

bipedalism. – Tiger (1979, p. 35)

Tiger’s quotation suggests that the trait of being optimistic or pessimistic has

biological origins as similarly stated by evolutionary hypotheses, which assume that

something genetic underlies the trait that is selected. Basically, evolutionary

psychology focuses on general traits, and provides interpretations for distal causes of

these traits relative to other species in terms of the environmental risks faced by the

species and of their physical properties in dealing with these challenges.

Optimism has had a profound influence in the fields of counselling,

psychology, and sociology. The psychological accounts of optimism have long been

involved in the pursuit of a more adaptive life for human beings. No matter what the

appoach in defining and measuring optimism, it has been widely accepted that being

optimistic represents the tendency and desire to maintain positive and adaptive

thinking, leading to positive emotions and behaviors, for promising expectations and

optimisitc attributions in life (Alarcon, Bowling, & Khazon, 2013; Andersson, 1996;

Carver & Scheier, 2014; Carver et al., 2010; Forgeard & Seligman, 2012).

Two main approaches of optimism, dispositional optimism and optimistic

explanatory style, were the core variables in my research of understanding optimism.

In a series of studies I investigated several aspects concerning these two traits,

including their psychometric structures, the relationship between dispositional

optimism and explanatory style, associations of optimism with psychological well-

being and personality, and potential cultural influences on optimism between two

ethnic groups. In addition, I conducted two pilot studies in the field of attributional



style, including the exploration of attributional style in others, and an examination of

the potential self-serving attributional bias in self- and other-settings. Finally and

importantly, after examining what optimism is and how we measure it, I explored the

possibility of optimism interventions on depressive symptoms.

Most of the studies involved Chinese undergraduate samples, except the cross-

cultural study of optimism. Findings in these studies are helpful to improve the

understanding of optimism in non-English speaking countries. In the first part of this

chapter, I reviewed and summarized the main findings concerning the psychometirc

structure of the basic measures in my study: the ASQ for explanatory style and the

LOT-R for dispostional optimism. Additionally, correlations between dimensions of

these two measures and two important psychogical variables, which include

psychological well-being and the FFM, were also briefly reported.

9.1 Summary of main findings

ASQ: three valence-independent cognitive styles

Explanatory style or attributional models of optimism, as measured by the ASQ,

focus on three aspects of attributions for the causes of positive and negative events:

internality, stability, and pervasiveness. These three aspects are assumed to cluster

within each valence forming explanatory-style factors and these in turn are predicted

to correlate negatively. Optimistic explanatory styles are associated with the belief

that the causes of negative events are external, unstable, and pervasive, while a

pessimistic attributional style assigns negative events as brief, affecting more than

one aspect of life, and internally caused (Forgeard & Seligman, 2012). However,

several empirical studies reported positive and negative events being uncorrelated

(Philip J. Corr & Jeffrey A. Gray, 1996; Peterson et al., 1982). With a non-Western

sample, I carried out the first test of the full structure of attributions controlling for

response non-independence.



Both negative and positive event attributions fit a three-dimensional structure

just as reported by Hewitt et al. (2004) and Higgins et al. (1999). However, the joint

modelling analysis of positive and negative events revealed that attributional biases

to positive and negative events were uncorrelated (see Figure 2.8). This model was

successfully replicated in an independent sample. Cognitive styles emerged as an

important influence on responding: valence-independent cognitive styles accounted

for 85 percent of variance in the latent-factor model. This suggests that subjects

apply consistent cognitive styles independent of event-valence, with personal

tendencies to explain events as, for instance, global or local independent of event

valence. Subjects rating negative events as global tended also to describe positive

events in terms of pervasive attributions, and likewise for the other two styles. In

conclusion, attributions may be best viewed as reflecting large differences in

cognitive style (independent of event valence), and smaller independent positive–

and negative-event biases.

LOT-R: separating dispositional optimism from dispositional pessimism

As the most frequently used measure of dispositional optimism, the LOT or its

revised version, the LOT-R, has been applied widely in numerous studies. Though

dispositional optimism was originally presumed to be a bipolar dimension, as

measured by the LOT or LOT-R (Scheier & Carver, 1985; Scheier et al., 1994), a

debate concerning the dimensionality of this variable has begun. More and more

evidence indicates that the LOT or LOT-R may reflect a two-factor model of

dispositional optimism. The positively and negatively phrased items in the measure

split into two factors, namely “optimism” and “pessimism”, representing two distinct

traits (Chang et al., 1997; L. Chang & McBrideChang, 1996; Creed et al., 2002;

Roysamb & Strype, 2002). Structural modelling of the LOT-R in my study

corresponded with previous findings that this measurement is better to be explained

as a two-dimensional structure scale.

Additionally, correlations between dispositional optimism and explanatory

style were examined. LOT-R optimism was positively correlated with ASQ Total

and ASQ Positive, but the correlation was lower than it has been reported by earlier



studies. Moreover, LOT-R optimism was positively correlated with Stable Positive

and negatively correlated with Stable Negative, but had no significant correlation

either with ASQ Negative or with any three dimensions of negative events. No

significant correlation was found between ASQ Pessimism and any ASQ dimensions.

Because only a general correlation between the LOT-R and ASQ composite has been

reported in most previous studies, results in this study provided at least some benefits

to better understanding the relationship between dispositional optimism and

explanatory style.

Furthermore, my study provided empirical evidence of the correlational

patterns between explanatory style and dispositional optimism in a non-Western

sample. The results were generally consistent with findings of previous research in

Western samples. That is, explanatory style and dispositional optimism are weakly

correlated (Forgeard & Seligman, 2012).

Optimism and the Five-Factor Model of personality

Optimism has been identified as thoughts and beliefs people hold for life and the

future. Both attributional style and dispositional optimism have been assessed largely

through their linkage to traditional personality traits, especially the FFM.

For explanatory style, attributions for negative events has been found to be

negatively correlated with Conscientiousness (Musgrave-Marquart et al., 1997).

Correlational analyses between ASQ and FFM dimensions in my study supported

this finding. Attributional styles for negative and positive events have been found to

have different correlational patterns with the FFM. While the ASQ Negative is

positively correlated with Neuroticism, and is negatively correlated with

Extraversion and Conscientiousness, ASQ Positive is positively related to four of the

five NEO-PI-R dimensions, excepting Neuroticism. Though attributions for positive

and negative events may reflect differentiated cognitive styles, these results suggest

that Conscientiousness may be considered as an important predictor of attributional

style.



For dispositional optimism, its bidimensional structure has been further

supported in an SEM model correlating LOT-R and the FFM. An initial base model

that incorporates two differentiable but related factors (LOT-R Optimism and LOT-R

Pessimism) through their links to the FFM was proposed and supported by data.

Based on these findings, dispositional optimism may be best viewed as reflecting two

distinct traits, which are reflected in LOT-R Optimism items and LOT-R Pessimism

items.

Additionally, associations among the LOT-R, ASQ, and NEO-PI-R scales

provide at least some evidence of the related but distinct relationship between the

two optimism structures. Though LOT-R Optimism and ASQ Positive both had

strong associations with the same four FFM factors, Neuroticism was only

significantly correlated with LOT-R Optimism but not ASQ Positive. In addition,

Openness only significantly correlated with LOT-R Pessimism but not with ASQ

Negative.

Mixed correlational patterns emerged when gender differences were taken

into account in analysing the relationship between optimism and personality. Results

showed that Agreeableness was the critical factor in differentiating attributional

styles of men and women. Specifically, Agreeableness was correlated with ASQ

Positive for men but not women, while it was correlated with ASQ Negative for

women but not men. For associations between LOT-R and NEO-PI-R scales, gender

differences presented a more complicated pattern. While Agreeableness was

correlated with dispositional pessimism for men but not for women, Openness was

correlated with both dispositional optimism and dispositional pessimism for women

but not for men.

Moreover, in the correlational analysis on optimism and specific facets of

each FFM factor, results demonstrated the positive correlations between optimism

(both high levels of dispositional optimism and optimistic explanatory styles) and

psychological well-being, such as lower depression scores and higher levels of

positive emotions.



Optimism: a strong predictor of psychological well-being

Dispositional optimism and explanatory style have been consistently related to health

and well-being. Previous investigations have shared two primary limitations. They

either have exclusively assessed only one construct of optimism (attributional style

or dispositional optimism) or merely measured one approach of well-being

(subjective well-being or psychological well-being). Even in studies where the two

fundamental constructs of optimism have both been assessed, the potential mediating

model linking all these constructs has not been examined. My study used SEM

models to construct relationships between optimism and psychological well-being.

Results from my study indicate that more optimistic individuals report a

higher level of psychological well-being, which is consistent with studies conducted

in Western participants. That is, individuals who have positive expectations for the

future are more likely to report high levels of psychological well-being. Optimistic

explanatory style may serve as another protective factor for well-being. There is

evidence that optimists tend to face adversity and deal with negative situations more

effectively than pessimists and can cope more adaptively with stress and, in turn,

gain more psychological benefits (Scheier & Carver, 1992).

Also, consistent with previous studies that individuals who have an optimistic

explanatory style are more likely to report higher levels of psychological well-being

than people with a pessimistic attributional style (Wise & Rosqvist, 2006), the

current results revealed that higher scores on ASQ Positive and lower scores on ASQ

Negative were significantly correlated with higher levels of psychological well-being

dimensions. Optimistic explanatory style may serve as a protective factor for well-

being.

The proposed mediating role of dispositional optimism between explanatory

style and psychological well-being was supported in the study. Results from

structural equation modelling indicated that explanatory style, dispositional optimism,

and PWB are positively associated with each other; dispositional optimism and

optimistic explanatory style are predictors of psychological well-being; and



dispositional optimism acts as a mediator between explanatory style and


Overall, this study provides consistent evidence of, and further support for,

the beneficial effects of both two types of optimism on psychological well-being in a

college student sample. Both dispositional optimism and optimistic explanatory style

are strong predictors of psychological well-being. Explanatory style and dispositional

optimism are weakly correlated (Forgeard & Seligman, 2012), though both

constructs are moderately correlated with well-being (Carver et al., 2010). Overall,

these findings are consistent with previous research in Western samples.

9.2 Does culture make a difference

Several studies investigated the universality of optimism using large sample sizes.

Fischer and Chalmers (2008) examined levels of dispositional optimism using a

meta-analytic approach, and reported that overall cultural differences in dispositional

optimism were small. The study involved a sample of 89,138 participants (more than

half American) from 22 countries. The optimism scores on average were found to be

significantly higher than the midpoint of LOT responses. Later, Gallagher et al.

(2013) examined the cross-cultural effects in optimism using a much larger sample

(n = 150,048) collected in the first wave of the Gallup World Poll involving

participants from 148 countries. They found that dispositional optimism was

significantly correlated with subjective well-being and perceived physical health both

at the country and the individual level, though the associations varied across

countries.

Cultural differences in optimism have been found in cross-cultural studies as

well. Michalos (1988) conducted one of the very first studies examining the

worldwide optimism level using the Gallup Report data. Participants from 31

countries were asked a single question: “So far as you are concerned, do you think

that 1987 will be better or worse than 1986?” Participants who gave the positive

answer to this question were classified as being optimistic for the future. Results



revealed that the optimism level of most countries and most individuals was not

promising, with only an average of 32 percent of participants in all countries

expecting a better future for the next year. The fact that some countries (such as

Canada and the U.S.) had a higher ratio of optimistic people than average indicated

potential cultural differences in optimism.

Still, in a meta-analytic study of the relationship between dispositional

optimism and coping style, Nes and Segerstrom (2006) reported that stronger

correlations between optimism and coping were found among participants in

English-speaking countries than their counterparts in non-English-speaking countries.

The results indicated that culture and language may have impacts on the optimism-

coping relationship.

The universality of the self-serving bias in causal explanations was supported

by the data in my study. Both ethnic groups (Mainland Chinese and White British)

reported positive ASQ Total scores, indicating a universal trend of holding an

optimistic explanatory style or a self-serving bias in causal attributions no matter

what the cultural background.

Admittedly, culture still plays a part in labelling different patterns and merits

of optimism, including both dispositional optimism and explanatory style. My study

concerning potential cultural differences on these two optimism approaches tested

several hypotheses. The first aim was to test whether similar psychometric structures

were applicable for the White British sample as in the Mainland Chinese sample. the

results revealed that a model of causal attributions for positive events in terms of

three correlated factors of globality, stability, and internality adequately accounted

for responses to these positive but not negative events in the ASQ. For the LOT-R

construct, a similar two-factor model of dispositional optimism was supported by my

study in the White British sample.

Results revealed several basic points concerning potential cultural differences

in optimism between the two ethnic groups. First, they were found to differ among a

number of important outcome variables in optimism. For example, Mainland Chinese



showed a more pessimistic explanatory style for explaining ASQ negative events

than their White British counterparts, which supported the proposal that Easterners

tend to use more unfavourable attributions for negative events than Westerners. For

explanations of ASQ positive events, unexpected patterns emerged. Mainland

Chinese expressed a more optimistic attributional style than White British in

attributing positive events, which was inconsistent with some previous research.

However, the results were consistent with our analysis that individuals tend to

produce similar patterns of explanations based on cognitive style rather than on event

type. These mixed results suggest that the cultural influence on optimism is not

uniform for at least some of the differentiated dimensions.

Additionally, the associational patterns between measuring scores of

optimism dimensions was quite similar for the two ethnic groups concerned, for

example, positive correlations between LOT-R-optimism and optimistic explanatory

style were found for both Mainland Chinese and White British. Discrepancies

between these two ethnic groups exist, however. For example, there was a weaker

negative association between LOT-R-optimism and LOT-R-pessimism for White

British than for Mainland Chinese, indicating a potential cultural or linguistic effect

on optimism measuring outcomes.

One aspect worth noting was the change in tendency of traditional

discrepancies in optimism between Easterners and Westerners found in my study.

The Mainland Chinese sample in this study expressed higher levels of LOT-R-

optimism and lower levels of LOT-R-pessimism than their White British

counterparts. In addition, Mainland Chinese also reported a more optimistic

explanatory style for positive events than White British. All these results are

inconsistent with traditional views of cultural discrepancies between the East and the

West. However, these findings are not as unexpected as they may seem, if two

factors are considered. First, it has been argued that broader social factors should be

taken into account in understanding optimism and pessimism (Lee & Seligman,

1997). Accordingly, these seemingly unexpected findings might be unique to this

young Chinese population. The relatively recent fast economic growth of China may

provide an explanation for Chinese people, especially as young generations feel more



optimistic and confident than previously, therefore dimming previous cultural

influences on optimism.

Secondly, as noted by some researchers, one of the major concerns in examing

culture differences in optimism is that it might be a problem for Easterners to get the

exact meaning of LOT-R items since this questionnaire has been developed on the

basis of Western cultures (Anderson, 1999). Hence, it is possible that there are slight

gaps in understanding the meaning of optimism and pessimism. At the very least, this

is in line with some results from previous research, as discussed earlier, that found no

group differences in optimism across cultures (Chang et al., 2003), or differences that

were more nuanced (Chang, 1996).

It should also be bear in mind that both these ethnic groups reported positive

ASQ Total scores in spite of differences in explanatory style between these two

cultural groups. This result indicated taht no matter what their cultural background

was, individuals tend to explain positive events with more internal, stable and global

causes than negative events. This conclusion is consistent with previous cross-

cultural evidence (e.g., Higgins & Bhatt, 2001), revealing that there is a universal

trend of positive bias in causal attributions.

9.3 Do people exhibit bias in attributing causes to events happening to others?

Though self-serving bias and self-versus other bias in causal attributions are

theoretically linked to each other, these two attributional biases have been studied

separately in prior literature. Unlike self-serving attributional bias that is mainly

assessed by the three-dimensional ASQ, self-versus other bias in causal attributions

has been restricted to the dimension of internality using diverse measures. To include

both self-serving bias and self-other bias in attributions into the widespread three-

dimensional model, I combined these two biases systematically across subjects (self

and other), valences (positive and negative events), and causes (traits and states) by



using the ASQ and a rewritten novel version of this measure (ASQ-Other), in which

participants generated attributions for events occurring to others.

Data and modelling analysis supported a model of causal attribution in terms of

three correlated factors of internality, stability, and globality accounting for

responses to both positive and negative events in the ASQ-Other, just as in the ASQ.

In particular, the ASQ-Other scale appears to be a valid and reliable measure, and

should be used in future studies to measure how people attribute others’ life events

outcomes.

The ASQ and the ASQ-Other were then used to assess self-serving attributional

bias and self-other attributional bias respectively. For self-serving attributional bias,

findings demonstrated that individuals tend to maximise positive and minimise

negative future outcomes in making attributions, thus show a self-protective bias in

causal explanations for personal outcomes or situations. This self-serving bias

manifested in each of the three attributional dimensions across event valence. When

individuals assign causal explanations for life events, they prefer giving more

internal, stable and pervasive causes for positive outcomes than for negative

outcomes. For unfavourable situations, individuals have the tendency of attributing

those situations to external, unstable, and specific causes.

For self-versus-other bias, results showed that people have more optimistic

explanatory styles for similar situations for themselves than for other people. This

self-versus-other bias exist in people’s attributions for both positive and negative

events. While individuals attribute others’ positive situations to external variables,

they explain their own positive outcomes using more favourable internal causes. The

opposite is true for negative situations. In summary, explanations for causes of

positive and negative events can be differentiated between self and other. Individuals

give more optimistic explanations for themselves than they did for others.

Additionally, results revealed that participants tend to attribute internal, stable,

and global attributions for positive events while they generate external, unstable, and

specific explanations for negative events no matter whether the subjects are



themselves or other people. Though people tend to have a more optimistic

explanatory style in events for themselves than for others, they expressed an

optimistic-biased attribution in explaining the causes of life events for other people.

Since prior studies in attributional bias have mainly been conducted in

Westerners, results confirming the existence of two forms of attributional biases in

an Eastern sample provided further evidence to prior findings. It appears that there

may be a universal tendency for individuals to protect themselves against negative

feelings by using an optimistic attributional style.

In summary, the results show that consistent with prior studies, these two

cognitive biases in causal attribution, or a tendency to hold an optimistic explanatory

style, also exist in at least the ,non-Western group in this study. Findings in the

current study demonstrated that causal attributions about life events possess a self-

protection feature, as suggested by Heider (1958). That is, individuals tend to

maximize positive and minimize negative future outcomes in making attributions,

thus showing a self-protective bias in causal explanations for personal outcomes or

situations.

9.4 Effective optimism interventions for depression

Due to diverse causes pf life transitions, such as challenges of living in a

different and unfamiliar environment, first-year undergraduate students have often

been found vulnerable to negative feelings, such as depression and anxiety, which

can negatively affect quality of life and academic performance (Brandy et al., 2015;

Negovan & Bagana, 2011). Previous studies have indicated that dispositional

optimism and attributional style may play an important role in psychological

adjustment during the first year in university (Brissette et al., 2002; Chemers et al.,

2001; Peterson & Barrett, 1987; Reisbig et al., 2012).

Previous research has shown that the effortful practice of imagining one’s best

possible future self and figuring out optimistic attributional styles for life events lead

to improved well-being and decreased negative feelings (Fresco et al., 2009;



Lyubomirsky et al., 2011; Peters et al., 2013). Although a number of studies have

explored the impact of BPS and SOT, no extensive study has tested their

effectiveness for treating depression has yet, to my knowledge, been conducted. I

applied these two forms of optimism interventions in two studies to evaluate the

feasibility of these interventions in depressed first-year college students. According

to previous findings concerning the influences of attributional style and dispositional

optimism have on academic performance, depression, and psychological adjustment,

the current investigations aimed to evaluate the feasibility of a prophylactic optimism

intervention in reducing depressive symptoms and improving psychological well-

being. Specifically, I sought to examine the beneficial effects of practicing SOT and

BPS daily on depressive symptoms, subjective well-being, dispositional optimism,

and explanatory style in a non-Western population.

The first pilot study combined an individual counselling session and self-

administered optimism manipulations to investigate the potential benefits of

optimism intervention. Results showed that individuals in the experimental condition

were less depressed than those in the control condition at post-intervention and two

follow-ups. Study 1 also showed that optimism interventions were beneficial in

developing optimistic explanatory styles, especially for attributions for negative

events. Extending previous findings that imagining and writing about a BPS leads to

an decrease in negative feelings (Shapira & Mongrain, 2010) and that practicing

optimistic attributions results in a reduction of depressive symptoms (Fresco et al.,

2009), the first study showed that daily practice of BPS an SOT for two weeks can

lead to sustained decrease in depression. In comparison with participants of the

control group, results revealed that individuals who practiced the BPS and SOT

techniques experienced less depressive symptoms and generated more optimistic

explanatory styles.

Though Study 1 has demonstrated that supervised and self-monitored optimism

interventions results in greater decreases in depressive symptoms in the experimental

condition, it raised the concern that this beneficial effect might be due to the

individual face-to-face counselling sessions in the interventions. To test whether self-



directed and self-administered optimism interventions could result in similar benefits

for decreasing depression, a second study was conducted.

The second study demonstrated that minimally supervised and self-administered

optimism interventions for a two-week period could result in decreases in depressive

symptoms and pessimistic explanatory styles and enhance dispositional optimism.

Although participants in the experiment group did not show a significantly greater

decrease in dispositional pessimism or a significantly greater increase in subjective

well-being, findings indicated that increases in dispositional optimism and decreases

in pessimistic explanatory style were associated with decreases in depressive

symptoms. Moreover, the benefits in decreased depression in the intervention group

was continued one month and three months after the intervention. the results

indicated that a brief and self-monitoring intervention is effective in reducing

symptoms of depression and enhancing well-being.

In general, both studies found evidence that a best possible self (BPS) imagery

intervention and self-administered optimism training in attributional style (SOT)

reduces the incidence of episodes of mild-to-moderate depression compared to a

control condition.

9.5 Deeper understanding of optimism: theoretical contributions to optimism literature and future directions

Although it is still not clear what the exact relationship between explanatory style

and dispositional optimism is, findings from the literature are mostly consistent.

Attributional style is related to a variety of psychological and physical health indices,

including academic achievement, depression, and physical illness (Wise & Rosqvist,

2006). Peterson and Seligman (1984) reviewed a variety of evidence showing that a

pessimistic attributional style predicts increases in depression over time in different

populations, such as lower-class women, children, and depressed patients. Similarly,

dispositional optimists report fewer depressive symptoms and fewer physical health



problems than pessimistic people (Carver & Scheier, 2014; Carver et al., 2010;

Scheier & Carver, 1987, 1992). These associations between the tendency to maintain

positive expectations for the future and improved well-being have been widely

recognized (Gallagher et al., 2013). My studies have replicated findings of a positive

relationship between optimism and well-being.

Previous studies have tried to identify optimism within a broad personality

domain, and it has been suggested that optimism represents a blend of Neuroticism

and Extraversion (Marshall et al., 1992). However, later work tends to support the

view that optimism also has some overlap with other Big Five Factors (Kam &

Meyer, 2012; Poropat, 2002; Sharpe et al., 2011). Findings in my study also support

this view.

In summary, optimism is a personality trait that can be related to nearly every

aspect of people’s life. It is clear that for encouraging people in general to be more

hopeful about the future, optimism interventions related to both attributional style

and dispositional optimism are worth further exploration. Though this stage of

research is focused on several aspects of feasibility, such as manual development,

pilot testing, and psychometric evaluation, the current investigation in my studies

supports the feasibility of prophylactic optimism intervention in reducing depressive

symptoms. The results indicate that positive interventions using optimism may be

suitable to study and establish effective early intervention for decreasing depressive

symptoms.

In recent years, the effects of positive thinking and behaviour have received

growing attention by psychologists, sociologists, anthropologists, clinicians, and

health professionals. With the increase in popularity of positive psychology,

optimism has gained more attention from the field of positive social science, and

allows for an examination of more aspects in life outcomes, such as the domain of

social relationships. It has been reported that optimism is linked to greater social

network size, and greater social support than pessimism (Carver & Scheier, 2014).

Given the accumulation of evidence, it is clear that optimism is an individual

difference variable that plays a central role in human experience in positive



psychology. Psychologists interested in optimism tend to correlate it with many other

psychological constructs, for instance those related to explanatory style and

dispositional optimism.

Although our findings provide some insight into the intricate covariations

frequently observed between certain psychological traits and optimism, a few

methodological and sampling limitations of my studies must be mentioned. First, all

the samples involved are consisted of college students, which might have specific

characteristics in optimism. Previous studies have shown that older people may have

different characteristics comparing with younger people. For example, in samples

including Americans and Hong Kong Mainland Chinese, You et al. (2009) reported

that older Mainland Chinese displayed a lower level of dispositional optimism than

did younger Mainland Chinese, whereas older Americans showed a higher level of

dispositional optimism than their younger counterparts. However, there is no

concrete evidence supporting this view in explanatory style in Chinese samples as far

as I know. Second, all the participants are undergraduates studying in the cities. The

level of optimism and correlations between optimism and other psychological

constructs, like psychological well-being, might vary to backgrounds of rural/urban

or different social economic status (Heinonen et al., 2006; MacLeod & Conway,

2005). Accordingly, further investigations and future studies would link optimism

variation to samples of several age groups, with different social backgrounds and

other features that might have influences on optimism. Third, it should be kept in

mind that SEM does not allow one to many any confident causal inferences about

relationships between variables. A model that fits the data well can only explain part

of the true correlations but not the whole truth. Thus, my conclusions got though

SEM modelling remain tentative. Additional work on these relations will strengthen

inferences regarding some pathways that have not been previously reported.

Fourth, though the two intervention studies have both supported effectiveness of

optimism intervention in promoting psychological well-being, especially in

decreasing depressive symptoms, they were only pilot studies with relatively small

samples of college students. It should be very cautious to generalize these findings in

people with wide backgrounds and varieties. Another possible limit was the use of



self-report surveys in assessing variables involved in all my studies. As discussed,

people’s self-reporting perceptions of optimism-related traits may be greatly affected

by social desirability. Data from multiple perspectives, such as reports from friends

and family members, might improve findings’ validity and reduce problems of

shared method variance.

9.6 Is optimism always good? Is pessimism always bad? The evolutionary explanations for optimism and pessimism

Having a different approach in dealing with people and happenings in the

surrounding world, in attempts to solve problems encountering in life, in attributional

styles to explain good or bad life events, in coping strategies facing difficult

situations, and even in attitudes dealing with social relationships, optimists and

pessimists behave differently in many core psychological and social processes, which

undoubtedly have substantial impacts on every aspect of their lives. Basically,

optimism and pessimism have been taken as inherent aspects of human nature and

also as individual differences in both theoretical discussions and empirical

investigations. Diverse benefits of optimism and concomitant drawbacks of

pessimism have been documented by a number of researches in psychology and

other social fields.

It has long been believed that positive thinking is linked to promising feeling.

Such an assertion has been examined over the last 35 years, with much solid

scientific evidence provided by psychologists through numerous empirical studies. In

addition to the benefits of being optimistic on physical health, it also has to be made

clear that positive thinking is linked to physical well-being only through a complex

process that involves intertwined biological, emotional, cognitive, and social

elements (Peterson & Bossio, 2001), but does not directly determine how well people

feel about their physical health.

The evidence reviewed in the prior sections suggests that being optimistic

seems like holding the keys to a rich and fulfilling life. Optimism is such an adaptive

feature that it is positively correlated with promising results in various contexts.



Conversely, pessimism is such an unfavourable trait that it indicates passivity,

failure, social estrangement, mortality, and depression.

Generally speaking, lines of research in optimism and pessimism are

surprisingly uniform, so much so that a popular trend of optimism has been created,

within psychology as well as the general public. Then one question comes: Why

pessimism has not been entirely abandoned in the life of human being? To answer

this question properly, we have to first entangle the relationship of optimism and

pessimism from an evolutionary view, and also review some concrete evidence of the

downside optimism and upside of pessimism.

Optimism has long been taken as an inherent aspect of human nature and one

of the most defining and adaptive characteristics of human being (Tiger, 1979). From

an evolutionary view, Tiger speculated that optimism first appeared when people

began to think about the future concerning dire consequences, which their own

mortality was included. To counteract the fear and powerlessness that these

anticipations might involve, something entailing hope had to be developed. Then

optimism came as an inherent and nature part of human nature.

To think about the evolutionary nature of optimism, we have to deal with the

relationship of optimism and pessimism. Are there effects of optimism above and

beyond those of the absence of pessimism? This intriguing question has to be

investigated first. Optimism and pessimism are usually taken as mutually exclusive,

but there is evidence that they are not. Taking one of the most popular measuring

tools of optimism, the LOT, as example, optimism was constructed reflecting a

bipolar construct (Scheier & Carver, 1985). That is, there is plentiful possibility that

some people expect both good things and bad things. Optimism and pessimism are

not exclusively independent of one another.

Similarly, though explanatory style was originally differentiated as two

independent categories, which assigns people an optimistic or a pessimistic

explanatory style. An optimistic explanatory style consists of explaining positive

events as enduring, global and internally generated, while also explaining negative

events as unstable, specific, and externally caused (Forgeard & Seligman, 2012).

Concept of attributional style also predicts that the three types of explanation are

correlated each other within at least within each event valence. Subsequent



researches have resolved in findings that are somewhat counterintuitive. For

instance, P.J. Corr and J.A. Gray (1996) investigated the factor structure of the ASQ

in two independent samples and found that positive and negative explanatory styles

were independent. The study of Bunce and Peterson (1997) also revealed that there is

no correlation between explanations for positive and negative events. This

independence was reported in my SEM analysis of ASQ in two Chinese samples as

well. Along these lines, as already noted, explanatory style derived from attributions

about negative events and explanatory style based on attributions about positive

events may be not as independent as originally thought. It might be best to view

explanatory style as a strategy of excuse making (Snyder et al., 1996). For most

individuals, mixed attributional styles should be expected: such as optimistic

explanations for negative events and pessimistic attributions for positive events.


heavy distinct behavioural consequences. For instance, low self-esteem is agreed to

be linked with internal attributions regarding negative events, while chronic

depression is suggested to result from stable attributions for negative events (Haugen

& Lund, 1998; Peterson et al., 1982). In this learned helplessness model, depression

emerges as a consequence of experience with uncontrollable negative events

(Abramson et al., 1978).

From the underlying assumption of positive psychology, psychological well-

being cannot be simply seen as the absence of distress and negative emotions.

Positive states or traits are not necessarily the obverse of negative experiences and

traits; and positive emotions and behaviours are described by a completely separate

psychological process that functions via an isolated neural mechanism (Duckworth et

al., 2005). Along these lines, dispositional optimism is not necessarily the obverse of

dispositional pessimism; and optimistic explanatory style is not exclusively absence

of pessimistic explanatory style.

In addition to the evolutionary explanations and theoretical origins of

optimism and pessimism, evidence from some empirical studies has proven that

optimism in some circumstances can have drawbacks and costs. Researchers have

begun to look for these qualifying conditions in various contexts. It is proposed that

optimists may have worse experiences in confronting negative outcomes than



pessimists due to their disconfirmed promising expectations (Gibson &

Sanbonmatsu, 2004). Accordingly, a question has then been raised virtually from the

inception of research on the optimism structure：Are there certain contexts or

situations in which optimism can potentially result in undesirable outcomes?

Some studies have tried to answer this question with concrete evidences. For

example, Gibson and Sanbonmatsu (2004) investigated relationships between

dispositional optimism and gambling expectations and behaviours. They reported

that optimists had more positive expectations for gambling than did pessimists, and

were more likely to maintain their betting even after poor outcomes. These findings

suggest that too much confidence and persistence might be counterproductive at least

in certain kinds of contexts, such as gambling.

Also, it has been suggested that optimism might not have the same protective

benefits as pessimism because optimists tend to see only what they want to see and

might ignore information of potential health threats (Norem & Chang, 2002). For

example, Luo and Isaacowitz (2007) examined how optimists process health-related

information regarding skin cancer. Their results indicated that pessimists paid more

attention to negative health-related information than optimists in certain kinds of

situations, though optimists were more likely to perform adaptive health-promoting

behaviours. These results suggest the possibility of different information-processing

methods between optimists and pessimists.

In another study, Hmieleski and Baron (2009) reported a negative relationship

between entrepreneurs’ optimism and their performance, defined as revenue and

employment growth of their new ventures. This negative relationship suggests that

optimists often hold unrealistic expectations and are overconfident, which was

assumed to lead to poor decision-making in processing negative information. In a

very recent study, Lau et al. (2014) did not find a positive relationship between

optimism and positive affect. Instead, pessimism showed beneficial effects on

positive affect and feelings of success when optimism and internal attribution were

disentangled.

Though these rare findings of potential adverse effects of optimism seem

small in comparison with the vast beneficial effects of being optimistic, they should

be taken into account when considering the effects of optimism, at least in certain



kinds of contexts and situations. It should be kept in mind that pessimism is an

independent trait that has its own evolutionary origin and theoretical meaning.

Optimism and pessimism are not the absence of each other.


258

Reference

Abdel-Khalek, A., & Lester, D. (2006). Optimism and pessimism in Kuwaiti and

American college students. International journal of social psychiatry, 52(2),

110-126.

Abramson, & Metalsky. (1989). The Cognitive Style Questionnaire: Measurement of

negative cognitive styles about self and consequences. Unpublished

manuscript.

Abramson, Metalsky, G. I., & Alloy, L. B. (1989). Hopelessness depression: A

theory-based subtype of depression. Psychological review, 96(2), 358.

Abramson, Seligman, & Teasdale. (1978). Learned helplessness in humans: critique

and reformulaiton. Journal of Abnormal Psychology, 87(1), 49-74. doi:

10.1037//0021-843x.87.1.49

Ahrens, A. H., & Haaga, D. A. F. (1993). The specificity of attributional style and

expectations to positive and negative affectivity, depression, and anxiety.

Cognitive therapy and research, 17(1), 83-98. doi: 10.1007/bf01172742

Alarcon, G. M., Bowling, N. A., & Khazon, S. (2013). Great expectations: A meta-

analytic examination of optimism and hope. Personality and Individual

Differences, 54(7), 821-827. doi: http://dx.doi.org/10.1016/j.paid.2012.12.004

Anderson, C. A. (1999). Attributional style, depression, and loneliness: A cross-

cultural comparison of American and Chinese students. Personality and

Social Psychology Bulletin, 25(4), 482-499. doi:

10.1177/0146167299025004007

Andersson, G. (1996). The benefits of optimism: A meta-analytic review of the Life

Orientation Test. Personality and Individual Differences, 21(5), 719-725.

Applebaum, A., Stein, E., Rosenfeld, B., & Breitbart, W. (2012). Optimism, Social

Support and Mental Health Outcomes in Patients With Advanced Cancer.

Psycho-Oncology, 21, 96-96.

Arbuckle, J. (2008). Amos 17.0. 0. AMOS Development Corporation, Crawfordville.

Armbruster, D., Pieper, L., Klotsche, J., & Hoyer, J. (2015). Predictions get tougher

in older individuals: a longitudinal study of optimism, pessimism and

depression. Social psychiatry and psychiatric epidemiology, 50(1), 153-163.

doi: 10.1007/s00127-014-0959-0

Ashkanasy, N. M. (1997). Attributions for the performance of self and other: it

matters who the “other” is. Australian journal of psychology, 49(1), 14-20.

Atienza, A. A., Stephens, M. A. P., & Townsend, A. L. (2004). Role stressors as

predictors of changes in womens’ optimistic expectations. Personality and

Individual Differences, 37(3), 471-484.

Augusto-Landa, J. M., Pulido-Martos, M., & Lopez-Zafra, E. (2011). Does Perceived

Emotional Intelligence and Optimism/pessimism Predict Psychological Well-

being? Journal of Happiness Studies, 1-12.

Austenfeld, J. L., Paolo, A. M., & Stanton, A. L. (2006). Effects of Writing About

Emotions Versus Goals on Psychological and Physical Health Among

Third‐Year Medical Students. Journal of Personality, 74(1), 267-286.

Austin, J. T., & Vancouver, J. B. (1996). Goal constructs in psychology: Structure,

process, and content. Psychological bulletin, 120(3), 338.

http://dx.doi.org/10.1016/j.paid.2012.12.004


259

Aydin, G. (1988). The remediation of children's helpless explanatory style and

related unpopularity. Cognitive therapy and research, 12(2), 155-165.

Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction

in social psychological research: Conceptual, strategic, and statistical

considerations. Journal of personality and social psychology, 51(6), 1173.

Beck. (1976). Cognitive therapy and the emotional disorders.

Beck, J. S., & Beck, A. T. (2011). Cognitive behavior therapy: Basics and beyond:

The Guilford Press.

Bennett, K. K., & Elliott, M. (2005). Pessimistic explanatory style and cardiac health:

What is the relation and the mechanism that links them? Basic and Applied

Social Psychology, 27(3), 239-248. doi: 10.1207/s15324834basp2703_5

Billingsley, K. D., Waehler, C. A., & Hardin, S. I. (1993). Stability of optimism and

choice of coping strategy. Perceptual and Motor Skills, 76(1), 91-97.

Blackwell, S. E., Rius-Ottenheim, N., Schulte-van Maaren, Y. W. M., Carlier, I. V.

E., Middelkoop, V. D., Zitman, F. G., . . . Giltay, E. J. (2013). Optimism and

mental imagery: A possible cognitive marker to promote well-being?

Psychiatry Research, 206(1), 56-61. doi: 10.1016/j.psychres.2012.09.047

Boker, S., Neale, M., Maes, H., Wilde, M., Spiegel, M., Brick, T., . . . Fox, J. (2011).

OpenMx: An Open Source Extended Structural Equation Modeling

Framework. Psychometrika, 76(2), 306-317. doi: 10.1007/s11336-010-9200-

6

Boker, S., Neale, M., Maes, H., Wilde, M., Spiegel, M., Brick, T. R., . . . Fox, J.

(2013). OpenMx: Multipurpose Software for Statistical Modeling (Version R

package version 1.3.2). Virginia. Retrieved from

http://openmx.psyc.virginia.edu.

Boland, A., & Cappeliez, P. (1997). Optimism and neuroticism as predictors of

coping and adaptation in older women. Personality and Individual

Differences, 22(6), 909-919.

Bradley, G. W. (1978). Self-serving biases in the attribution process: A

reexamination of the fact or fiction question. Journal of personality and

social psychology, 36(1), 56.

Brandy, J. M., Penckofer, S., Solari-Twadell, P. A., & Velsor-Friedrich, B. (2015).

Factors predictive of depression in first-year college students. Journal of

psychosocial nursing and mental health services, 53(2), 38-44. doi:

10.3928/02793695-20150126-03

Brissette, I., Scheier, M. F., & Carver, C. S. (2002). The role of optimism in social

network development, coping, and psychological adjustment during a life

transition. Journal of personality and social psychology, 82(1), 102.

Bryant, F. B., & Veroff, J. (2007). Savoring: A new model of positive experience.

Buchanan, G. M. C. E., & Seligman, M. E. P. (1995). Explanatory style: Lawrence

Erlbaum Associates, Inc.

Bunce, S. C., & Peterson. (1997). Gender differences in personality correlates of

explanatory style. Personality and Individual Differences, 23(4), 639-646. doi:

10.1016/s0191-8869(97)00072-x

Burns, & Seligman. (1989). Explanatory style across the life span: evidence for

stability over 52 years. Journal of personality and social psychology, 56(3),

471.

http://openmx.psyc.virginia.edu/


260

Burns, D. D. (1980). Feeling good: The new mood therapy. New York: William

Morrow.

Burton, C. M., & King, L. A. (2004). The health benefits of writing about intensely

positive experiences. Journal of Research in Personality, 38(2), 150-163.

Burton, C. M., & King, L. A. (2008). Effects of (very) brief writing on health: The

two‐minute miracle. British Journal of Health Psychology, 13(1), 9-14.

Cabanis, M., Pyka, M., Mehl, S., Müller, B. W., Loos-Jankowiak, S., Winterer,

G., . . . Rapp, A. M. (2013). The precuneus and the insula in self-attributional

processes. Cognitive, Affective, & Behavioral Neuroscience, 13(2), 330-345.

Cadinu, M. R., Arcuri, L., & Kodilja, R. (1993). Sefl-serving biases - the role of

perspective-taking. European Journal of Social Psychology, 23(1), 97-102.

doi: 10.1002/ejsp.2420230109

Campbell, D. T., & Fiske, D. W. (1959). Convergent and Discriminant Validation by

the Multitriat - Multimethod Matrix. Psychological bulletin, 56(2), 81-105.

doi: 10.1037/h0046016

Caprara, G. V., Fagnani, C., Alessandri, G., Steca, P., Gigantesco, A., Sforza, L. L.

C., & Stazi, M. A. (2009). Human optimal functioning: The genetics of

positive orientation towards self, life, and the future. Behav Genet, 39(3),

277-284.

Carver, & Gaines, J. G. (1987). Optimism, pessimism, and postpartum depression.

Cognitive therapy and research, 11(4), 449-462.

Carver, Pozo, C., Harris, S. D., Noriega, V., Scheier, Robinson, D. S., . . . Clark, K.

C. (1993). How coping mediates the effect of optimism on distress: a study of

women with early stage breast cancer. Journal of personality and social

psychology, 65(2), 375.

Carver, & Scheier. (2001). Optimism, pessimism, and self-regulation (chapter 2). In

E. C. Chang (Ed.), Optimism & Pessimism, Implications for theory, research,

and practice. Washington, DC: American Psychological Association.

Carver, & Scheier. (2014). Dispositional optimism. Trends in cognitive sciences.

Carver, Scheier, & Segerstrom. (2010). Optimism. Clinical Psychology Review,

30(7), 879-889.

Carver, Scheier, & Weintraub, J. K. (1989). Assessing coping strategies: a

theoretically based approach. Journal of personality and social psychology,

56(2), 267-283. doi: 10.1037//0022-3514.56.2.267

Cattell, R. B. (1943). The description of personality: Basic traits resolved into

clusters. Journal of Abnormal and Social Psychology, 38(4), 476-506. doi:

10.1037/h0054116

Cecil, M. A., & Medway, F. J. (1986). Attribution retraining with low-achieving and

learned helpless children. Techniques.

Chan, C.-M., & Tsoi, M. M. (1984). The BDI and stimulus determinants of

cognitive-related depression among Chinese college students. Cognitive

therapy and research, 8(5), 501-507.

Chang. (1996). Cultural differences in optimism, pessimism, and coping: Predictors

of subsequent adjustment in Asian American and Caucasian American

college students. Journal of Counseling Psychology, 43(1), 113.


261

Chang. (2001a). Introduction: Optimism and pessimism and moving beyond the most

fundamental question. Optimism and pessimism. Implications for theory,

research and practice, 4-12.

Chang. (2001b). Optimism & pessimism: Implications for theory, research, and

practice: American Psychological Association.

Chang, Maydeu-Olivares, A., & D'Zurilla, T. J. (1997). Optimism and pessimism as

partially independent constructs: Relationship to positive and negative

affectivity and psychological well-being. Personality and Individual

Differences, 23(3), 433-440.

Chang, Sanna, L. J., & Yang, K. M. (2003). Optimism, pessimism, affectivity, and

psychological adjustment in US and Korea: A test of a mediation model.

Personality and Individual Differences, 34(7), 1195-1208.

Chang, L., & McBrideChang, C. (1996). The factor structure of the Life Orientation

Test. Educational and Psychological Measurement, 56(2), 325-329. doi:

10.1177/0013164496056002013

Chemers, M. M., Hu, L.-t., & Garcia, B. F. (2001). Academic self-efficacy and first

year college student performance and adjustment. Journal of educational


Chen, L., & Zhang, J. (2004). General Life Satisfaction and Self-esteem of

University Students. Chinese Mental Health Journal, 4.

Cheng, S. T., & Hamid, P. N. (1997). Dispositional optimism in Chinese people:

What does the Life Orientation Test measure? International Journal of

Psychology, 32(1), 15-22. doi: 10.1080/002075997400935

Chirivella, E. C., Checa, I., & Budzynska, N. (2013). Coping, optimism and

satisfaction with life among Spanish and Polish football players: a

preliminary study. Revista De Psicologia Del Deporte, 22(2), 337-343.

Colby, D. A., & Shifren, K. (2013). Optimism, mental health, and quality of life: A

study among breast cancer patients. Psychology Health & Medicine, 18(1),

10-20. doi: 10.1080/13548506.2012.686619

Corr, P. J., & Gray, J. A. (1996). Structure and validity of the Attributional Style

Questionnaire: A cross-sample comparison. Journal of Psychology, 130(6),

645-657.

Corr, P. J., & Gray, J. A. (1996). Structure and validity of the Attributional Style

Questionnaire: A cross-sample comparison. The Journal of psychology,

130(6), 645-657.

Costa, & McCrae, R. R. (1992). The Revised NEO Personality Inventory (NEO-PI-R)

and NEO Five Factor Inventory (NEO-FFI) professional manual. .

Costa, & McCrae, R. R. (1995). Domains and facets: Hierarchical personality

assessment using the Revised NEO Personality Inventory. Journal of

personality assessment, 64(1), 21-50.

Creed, P. A., Patton, W., & Bartrum, D. (2002). Multidimensional properties of the

LOT-R: Effects of optimism and pessimism on career and well-being related

variables in adolescents. Journal of Career Assessment, 10(1), 42-61.

Csillik, A., Aguerre, C., & Bay, M. (2012). Positive psychotherapy of depression:

Specificities and clinical contributions. Annales Medico-Psychologiques,

170(8), 541-546. doi: 10.1016/j.amp.2011.10.012


262

Cuijpers, P., Smit, F., Bohlmeijer, E., Hollon, S. D., & Andersson, G. (2010).

Efficacy of cognitive–behavioural therapy and other psychological treatments

for adult depression: meta-analytic study of publication bias. The British

Journal of Psychiatry, 196(3), 173-178.

Cuijpers, P., Van Straten, A., Van Schaik, A., & Andersson, G. (2009).

Psychological treatment of depression in primary care: a meta-analysis. The

British Journal of General Practice, 59(559), e51.

Daukantaite, D., & Zukauskiene, R. (2012). Optimism and Subjective Well-Being:

Affectivity Plays a Secondary Role in the Relationship Between Optimism

and Global Life Satisfaction in the Middle-Aged Women. Longitudinal and

Cross-Cultural Findings. Journal of Happiness Studies, 13(1), 1-16. doi:

10.1007/s10902-010-9246-2

Derogatis, L. R., & Cleary, P. A. (1977). Confirmation of the dimensional structure

of the SCL‐90: A study in construct validation. Journal of Clinical

Psychology, 33(4), 981-989.

Derogatis, L. R., S, L. R., Covi, L., & Rickeis, K. (1973). The SCL-90: An outpatient

psychiatric rating scale. Psychopharmacology Bulletin, 9, 13-28.

Descartes, R. (1985). The philosophical writings of Descartes (Vol. 2): Cambridge

University Press (Original work published 1649).

Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with

life scale. Journal of personality assessment, 49(1), 71-75.

Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being:

Three decades of progress. Psychological bulletin, 125(2), 276.

Domino, B., & Conway, D. W. (2001). Optimism and pessimism from a historical

perspective. In E. C. Chang (Ed.), Optimism & pessimism: Implications for

theory, research, and practice (pp. 13-30). Washington, DC: American

Psychological Associatioin.

Duckworth, A. L., Steen, T. A., & Seligman. (2005). Positive psychology in clinical

practice. Annu. Rev. Clin. Psychol., 1, 629-651.

Ebert, S. A., Tucker, D., & Roth, D. (2002). Psychological resistance factors as

predictors of general health status and physical symptom reporting.

Psychology, health & medicine, 7(3), 363-375.

Eichner, K. V., Kwon, P., & Marcus, D. K. (2014). Optimists or optimistic? A

taxometric study of optimism. Psychological Assessment, 26(3), 1056-1061.

doi: 10.1037/pas0000010

Eysenck, H. J. (1965). A 3-factor theory of reminiscence. British Journal of

Psychology, 56(2-3), 163-181.

Ferguson, S. J., & Goodwin, A. D. (2010). Optimism and Well-Being in Older

Adults: The Mediating Role of Social Support and Perceived Control. The

International Journal of Aging and Human Development, 71(1), 43-68.

Fischer, R., & Chalmers, A. (2008). Is optimism universal? A meta-analytical

investigation of optimism levels across 22 nations. Personality and Individual

Differences, 45(5), 378-382. doi: http://dx.doi.org/10.1016/j.paid.2008.05.008

Folkman, S., & Lazarus, R. S. (1980). An analysis of coping in a middle-aged

commmunity sample. Journal of Health and Social Behavior, 21(3), 219-239.

doi: 10.2307/2136617

http://dx.doi.org/10.1016/j.paid.2008.05.008


263

Folkman, S., & Moskowitz, J. T. (2004). Coping: Pitfalls and promise. Annual

review of psychology, 55, 745-774. doi:

10.1146/annurev.psych.55.090902.141456

Fontaine, K. R., Manstead, A. S. R., & Wagner, H. (1993). Optimism, perceived

control over stress, and coping. European Journal of Personality, 7(4), 267-

281. doi: 10.1002/per.2410070407

Forgeard, M. J. C., & Seligman, M. E. P. (2012). Seeing the glass half full: A review

of the causes and consequences of optimism. Pratiques Psychologiques,

18(2), 107-120. doi: 10.1016/j.prps.2012.02.002

Fosnaugh, J., Geers, A. L., & Wellman, J. A. (2009). Giving Off a Rosy Glow: The

Manipulation of an Optimistic Orientation. The Journal of social psychology,

149(3), 349-364.

Fowler, J. W., & Peterson, P. L. (1981). Increasing reading persistence and altering

attributional style of learned helpless children. Journal of educational


Fresco, Alloy, L. B., & Reilly-Harrington, N. (2006). Association of attributional

style for negative and positive events and the occurrence of life events with

depression and anxiety. Journal of Social and Clinical Psychology, 25(10),

1140-1159. doi: 10.1521/jscp.2006.25.10.1140

Fresco, Moore, M. T., Walt, L., & Craighead, L. W. (2009). Self-administered

optimism training: mechanisms of change in a minimally supervised

psychoeducational intervention. Journal of Cognitive Psychotherapy, 23(4),

350-367.

Freud, S. (1961). The future of an illusion. In J. S. E. a. Trans.) (Ed.), The standard

edition of the complete psychological works of Sigmund Freud (Vol. Vol. 21).

London: Hogarth Press (Original work pulished 1927).

Fry, P. S. (1995). Perfectionism, humor, and optimism as moderators of health

outcomes and determinants of coping styles of women executives. Genetic

Social and General Psychology Monographs, 121(2), 211-&.

Gallagher, M. W., Lopez, S. J., & Pressman, S. D. (2013). Optimism Is Universal:

Exploring the Presence and Benefits of Optimism in a Representative Sample

of the World. Journal of Personality, 81(5), 429-440. doi:

10.1111/jopy.12026

Gibran, K. (1951). The Kahlil Gibran reader: inspirational writings. . New York:

Citadel Press.

Gibson, B., & Sanbonmatsu, D. M. (2004). Optimism, pessimism, and gambling:

The downside of optimism. Personality and Social Psychology Bulletin,

30(2), 149-160.

Gillham, & Reivich, K. (2004). Cultivating optimism in childhood and adolescence.

The Annals of the American Academy of Political and Social Science, 591(1),

146-163.

Gillham, Reivich, K. J., Freres, D. R., Chaplin, T. M., Shatté, A. J., Samuels, B., . . .

Gallop, R. (2007). School-based prevention of depressive symptoms: A

randomized controlled study of the effectiveness and specificity of the Penn

Resiliency Program. Journal of Consulting and Clinical Psychology, 75(1), 9.


264

Gillham, Reivich, K. J., Jaycox, L. H., & Seligman, M. E. (1995). Prevention of

depressive symptoms in schoolchildren: Two-year follow-up.

PSYCHOLOGICAL SCIENCE-CAMBRIDGE-, 6, 343-351.

Gillham, J. E., Shatté, A. J., Reivich, K. J., & Seligman, M. E. P. (2001). Optimism,

pessimism, and explanatory style. In E. C. Chang (Ed.), Optimism &

pessimism: Implications for theory, research, and practice (pp. 53-75).

Washington, DC, US: American Psychological Association.

Gillham, J. E., Tassoni, C. J., Engel, R. A., DeRubeis, R. J., & Seligman. (1998). The

relationship of explanatory style to other depression relevant constructs.

Unpublished manuscript. University of Pennsylvania.

Giltay, E. J., Geleijnse, J. M., Zitman, F. G., Hoekstra, T., & Schouten, E. G. (2004).

Dispositional optimism and all-cause and cardiovascular mortality in a

prospective cohort of elderly dutch men and women. Archives of general

psychiatry, 61(11), 1126-1135. doi: 10.1001/archpsyc.61.11.1126

Goodin, B. R., & Bulls, H. W. (2013). Optimism and the Experience of Pain:

Benefits of Seeing the Glass as Half Full. Current Pain and Headache

Reports, 17(5). doi: 10.1007/s11916-013-0329-8

Gordon, R. A. (2008). Attributional style and athletic performance: Strategic

optimism and defensive pessimism. Psychology of Sport and Exercise, 9(3),

336-350. doi: 10.1016/j.psychsport.2007.04.007

Gould, D., Dieffenbach, K., & Moffett, A. (2002). Psychological characteristics and

their development in Olympic champions. Journal of applied sport

psychology, 14(3), 172-204.

Graham, C. D., Weinman, J., Sadjadi, R., Chalder, T., Petty, R., Hanna, M. G., . . .

Rose, M. R. (2014). A multicentre postal survey investigating the

contribution of illness perceptions, coping and optimism to quality of life and

mood in adults with muscle disease. Clinical Rehabilitation, 28(5), 508-519.

Greenberg, J., Pyszczynski, T., Burling, J., & Tibbs, K. (1992). Depression, self-

focused attention, and the self-serving attributional bias. Personality and

Individual Differences, 13(9), 959-965.

Greenberg, J., Pyszczynski, T., & Solomon, S. (1982). The self-serving attributional

bias: Beyond self-presentation. Journal of Experimental Social Psychology,

18(1), 56-67.

Haeffel, G. J., Gibb, B. E., Metalsky, G. I., Alloy, L. B., Abramson, L. Y., Hankin, B.

L., . . . Swendsen, J. D. (2008). Measuring cognitive vulnerability to

depression: Development and validation of the cognitive style questionnaire.

Clinical Psychology Review, 28(5), 824-836.

Haeffel, G. J., & Vargas, I. (2011). Resilience to depressive symptoms: The

buffering effects of enhancing cognitive style and positive life events.

Journal of Behavior Therapy and Experimental Psychiatry, 42(1), 13-18.

Hanrahan, S. J., Grove, J. R., & Hattie, J. A. (1989). Development of a questionnaire

measure of sport-related attributional style. International Journal of Sport

Psychology, 20(2), 114-134.

Harrist, S., Carlozzi, B. L., McGovern, A. R., & Harrist, A. W. (2007). Benefits of

expressive writing and expressive talking about life goals. Journal of

Research in Personality, 41(4), 923-930.


265

Haugen, R., & Lund, T. (1998). Attributional style and its relation to other

personality dispositions. British Journal of Educational Psychology, 68, 537-

549.

Haynes, Perry, R. P., Stupnisky, R. H., & Daniels, L. M. (2009). A review of

attributional retraining treatments: Fostering engagement and persistence in

vulnerable college students Higher education: Handbook of theory and

research (pp. 227-272): Springer.

Haynes, Ruthig, J. C., Perry, R. P., Stupnisky, R. H., & Hall, N. C. (2006). Reducing

the academic risks of over-optimism: The longitudinal effects of attributional

retraining on cognition and achievement. Research in Higher Education,

47(7), 755-779.

Haynes, P. L., Ancoli-Israel, S., Walter, C. M., & McQuaid, J. R. (2012).

Preliminary Evidence for a Relationship Between Sleep Disturbance and

Global Attributional Style in Depression. Cognitive therapy and research,

36(2), 140-148. doi: 10.1007/s10608-011-9416-5

Heider, F. (1958). The psychology of interpersonal relations. New York: Wildy.

Heine, S. J., & Lehman, D. R. (1995). Cultural variation in unrealistic optimism:

Does the West feel more vulnerable than the East? Journal of personality and


Heine, S. J., Lehman, D. R., Markus, H. R., & Kitayama, S. (1999). Is there a

universal need for positive self-regard? Psychological review, 106(4), 766.

Heinonen, K., Raikkonen, K., Matthews, K. A., Scheier, M. F., Raitakari, O. T.,

Pulkki, L., & Keltikangas-Jarvinen, L. (2006). Socioeconomic status in

childhood and adulthood: Associations with dispositional optimism and

pessimism over a 21-year follow-up. Journal of Personality, 74(4), 1111-

1126. doi: 10.1111/j.1467-6494.2006.00404.x

Herzberg, P. Y., Glaesmer, H., & Hoyer, J. (2006). Separating optimism and

pessimism: A robust psychometric analysis of the Revised Life Orientation

Test (LOT-R). Psychological Assessment, 18(4), 433.

Hewitt, A. K., Foxcroft, D. R., & MacDonald, J. (2004). Multitrait-multimethod

confirmatory factor analysis of the attributional style questionnaire.

Personality and Individual Differences, 37(7), 1483-1491. doi:

10.1016/j.paid.2004.02.005

Higgins, N. C., & Bhatt, G. (2001). Culture moderates the self-serving bias: Etic and

emic features of causal attributions in India and in Canada. Social Behavior

and Personality: an international journal, 29(1), 49-61.

Higgins, N. C., Zumbo, B. D., & Hay, J. L. (1999). Construct validity of attributional

style: Modeling context-dependent item sets in the Attributional Style

Questionnaire. Educational and Psychological Measurement, 59(5), 804-820.

doi: 10.1177/00131649921970152

Hiroto, D. S., & Seligman, M. E. (1975). Generality of learned helplessness in man.

Journal of personality and social psychology, 31(2), 311.

Hjelle, L., Belongia, C., & Nesser, J. (1996). Psychometric properties of the life

orientation test and attributional style questionnaire. Psychological Reports,

78(2), 507-515.


266

Hmieleski, K. M., & Baron, R. A. (2009). Entrepreneurs' optimism and new venture

performance: A social cognitive perspective. Academy of management

Journal, 52(3), 473-488.

Ho, M. Y., Cheung, F. M., & Cheung, S. F. (2010). The role of meaning in life and

optimism in promoting well-being. Personality and Individual Differences,

48(5), 658-663.

Hochhausen, N., Altmaier, E. M., McQuellon, R., Davies, S. M., Papadopolous, E.,

Carter, S., & Henslee-Downey, J. (2007). Social support, optimism, and self-

efficacy predict physical and emotional well-being after bone marrow

transplantation. Journal of Psychosocial Oncology, 25(1), 87-101. doi:

10.1300/J077v25n01_05

Horney, D. J., Smith, H. E., McGurk, M., Weinman, J., Herold, J., Altman, K., &

Llewellyn, C. D. (2011). Association between quality of life, coping styles,

optimism, and anxiety and depression in pretreatment patient with head and

neck cancer. Head and Neck-Journal for the Sciences and Specialties of the

Head and Neck, 33(1), 65-71. doi: 10.1002/hed.21407

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure

analysis: Coventional criteria versus new alternatives. Structural Equation

Modeling, 6, 1-55.

Iacoviello, B. M., Alloy, L. B., Abramson, L. Y., Whitehouse, W. G., & Hogan, M. E.

(2006). The course of depression in individuals at high and low cognitive risk

for depression: A prospective study. Journal of affective disorders, 93(1–3),

61-69. doi: http://dx.doi.org/10.1016/j.jad.2006.02.012

Isaacowitz, D. M. (2005). Correlates of well-being in adulthood and old age: A tale

of two optimisms. Journal of Research in Personality, 39(2), 224-244.

Isaacowitz, D. M., & Seligman. (2002). Cognitive style predictors of affect change in

older adults. The International Journal of Aging and Human Development,

54(3), 233-253.

James, W. (1985). The varieties of religious experience (Vol. 13): Harvard

University Press.

Jaycox, L. H., Reivich, K. J., Gillham, J., & Seligman, M. E. (1994). Prevention of

depressive symptoms in school children. Behaviour Research and Therapy,

32(8), 801-816.

Johnstone, J., Rooney, R. M., Hassan, S., & Kane, R. T. (2014). Prevention of

depression and anxiety symptoms in adolescents: 42 and 54 months follow-

up of the Aussie Optimism Program-Positive Thinking Skills. Frontiers in

Psychology, 5. doi: 10.3389/fpsyg.2014.00364

Joiner, T. E., & Wagner, K. D. (1995). Attributional style and depression in children

and adolescents: A meta-analytic review. Clinical Psychology Review, 15(8),

777-798. doi: 10.1016/0272-7358(95)00046-1

Jones, E. E., & Nisbett, R. E. (1972). The actor and the observer: Divergent

perceptions of the causes of behavior. In E. E. Jones, D. E. Kanouse, H. H.

Kelley, R. E. Nisbett, S. Valins & B. Weiner (Eds.), Attribution: Perceiving

the causes of behavior (pp. 79-84): General Learning Press Morristown, NJ.

Jowsey, S. G., Cutshall, S. M., Colligan, R. C., Stevens, S. R., Kremers, W. K.,

Vasquez, A. R., . . . McGregor, C. G. A. (2012). Seligman's theory of

attributional style: optimism, pessimism, and quality of life after heart

http://dx.doi.org/10.1016/j.jad.2006.02.012


267

transplant. Progress in Transplantation, 22(1), 49-55. doi:

10.7182/pit2012451

Ju, H., Shin, J. W., Kim, C.-W., Hyun, M.-H., & Park, J.-W. (2013). Mediational

effect of meaning in life on the relationship between optimism and well-being

in community elderly. Archives of gerontology and geriatrics, 56(2), 309-313.

doi: 10.1016/j.archger.2012.08.008

Kam, C., & Meyer, J. P. (2012). Do optimism and pessimism have different

relationships with personality dimensions? A re-examination. Personality and

Individual Differences, 52(2), 123-127. doi: 10.1016/j.paid.2011.09.011

Kamen, L. P., & Seligman, M. E. P. (1987). Explanaotry style and health. Current

Psychology-Research & Reviews, 6(3), 207-218. doi: 10.1007/bf02686648

Kashima, Y., & Triandis, H. C. (1986). The Sefl-serving Bias in attibutions as a

coping stratege: A cross-cultural study. Journal of Cross-Cultural Psychology,

17(1), 83-97. doi: 10.1177/0022002186017001006

Kaslow, N. J., Tannenbau, R. L., & Seligman. (1978). The KASTAN: A children's

attributional style questionnaire. . Unpublished manuscript. University of

Pennsylvania. Philadelphia.

King, L. A. (2001). The health benefits of writing about life goals. Personality and

Social Psychology Bulletin, 27(7), 798-807.

Kirsch, I., Moore, T. J., Scoboria, A., & Nicholls, S. S. (2002). The emperor's new

drugs: An analysis of antidepressant medication data submitted to the US

Food and Drug Administration. Prevention & Treatment, 5(1), 23a.

Kleiman, E. M., Liu, R. T., & Riskind, J. H. (2013). Enhancing attributional style as

a resiliency factor in depressogenic stress generation. Anxiety, stress, and

coping, 26(4), 467-474. doi: 10.1080/10615806.2012.684381

Klein, D. C., Fencil-Morse, E., & Seligman, M. E. (1976). Learned helplessness,

depression, and the attribution of failure. Journal of personality and social


Kubzansky, L. D., Kubzansky, P. E., & Maselko, J. (2004). Optimism and pessimism

in the context of health: bipolar opposites or separate constructs? Personality

and Social Psychology Bulletin, 30(8), 943.

Kurtovic, A. (2012). Relationship between Attributions for Success and Failure and

Self-esteem, Hopelessness and Depression in Secondary School Students.

Croatian Journal of Education-Hrvatski Casopis Za Odgoj I Obrazovanje,

14(4), 771-797.

Lai, J. C. L. (1997). Relative predictive power of the optimism versus the pessimism

index of a Chinese version of the life orientation test. Psychological Record,

47(3), 399-410.

Lai, J. C. L., Cheung, H., Lee, W. M., & Yu, H. (1998). The utility of the revised life

orientation test to measure optimism among Hong Kong Chinese.

International Journal of Psychology, 33(1), 45-56.

Lai, J. C. L., & Yue, X. D. (2000). Measuring optimism in Hong Kong and mainland

Chinese with the revised Life Orientation Test. Personality and Individual

Differences, 28(4), 781-796. doi: 10.1016/s0191-8869(99)00138-5

Lau, S., Kubiak, T., Burchert, S., Goering, M., Oberlaender, N., von Mauschwitz,

H., . . . Hiemisch, A. (2014). Disentangling the effects of optimism and


268

attributions on feelings of success. Personality and Individual Differences, 56,

78-82. doi: 10.1016/j.paid.2013.08.030

Layous, K., Chancellor, J., Lyubomirsky, S., Wang, L., & Doraiswamy, P. M. (2011).

Delivering happiness: Translating positive psychology intervention research

for treating major and minor depressive disorders. The Journal of Alternative

and Complementary Medicine, 17(8), 675-683.

Ledrich, J., & Gana, K. (2013). Relationship between attributional style, perceived

control, self-esteem, and depressive mood in a nonclinical sample: A

structural equation-modelling approach. Psychology and Psychotherapy-

Theory Research and Practice, 86(4), 413-430. doi: 10.1111/j.2044-

8341.2012.02067.x

Lee, Y.-T., & Seligman, M. E. (1997). Are Americans more optimistic than the

Chinese? Personality and Social Psychology Bulletin, 23, 32-40.

Leibniz, G. W. (2010). Theodicy: Cosimo, Inc. (Original work published 1710).

Lemola, S., Räikkönen, K., Scheier, M. F., Matthews, K. A., Pesonen, A. k.,

Heinonen, K., . . . Kajantie, E. (2011). Sleep quantity, quality and optimism

in children. Journal of sleep research, 20(1pt1), 12-20.

Lemola, S., Raikkonen, K., Gomez, V., & Allemand, M. (2013). Optimism and Self-

Esteem Are Related to Sleep. Results from a Large Community-Based

Sample. International Journal of Behavioral Medicine, 20(4), 567-571. doi:

10.1007/s12529-012-9272-z

Li, C.-H. (2012). Validation of the Chinese Version of the Life Orientation Test With

A Robust Weighted Least Squares Approach. Psychological Assessment,

24(3), 770-776. doi: 10.1037/a0026612

Libby, R., & Rennekamp, K. (2012). Self-Serving Attribution Bias, Overconfidence,

and the Issuance of Management Forecasts. Journal of Accounting Research,

50(1), 197-231. doi: 10.1111/j.1475-679X.2011.00430.x

Linley, P. A., Nielsen, K. M., Wood, A. M., Gillett, R., & Biswas-Diener, R. (2010).

Using signature strengths in pursuit of goals: Effects on goal progress, need

satisfaction, and well-being, and implications for coaching psychologists.

International Coaching Psychology Review, 5(1), 6-15.

Littman-Ovadia, H., & Nir, D. (2014). Looking forward to tomorrow: The buffering

effect of a daily optimism intervention. Journal of Positive Psychology, 9(2),

122-136. doi: 10.1080/17439760.2013.853202

Llewellyn, C. D., Horney, D. J., McGurk, M., Weinman, J., Herold, J., Altman, K.,

& Smith, H. E. (2013). Assessing the psychological predictors of benefit

finding in patients with head and neck cancer. Psycho-Oncology, 22(1), 97-

105. doi: 10.1002/pon.2065

Lounsbury, J. W., Saudargas, R. A., & Gibson, L. W. (2004). An investigation of

personality traits in relation to intention to withdraw from college. Journal of

College Student Development, 45(5), 517-534.

Low, C. A., Stanton, A. L., & Danoff-Burg, S. (2006). Expressive disclosure and

benefit finding among breast cancer patients: mechanisms for positive health

effects. Health Psychology, 25(2), 181.

Lucas, R. E., Diener, E., & Suh, E. (1996). Discriminant validity of well-being

measures. Journal of personality and social psychology, 71(3), 616.


269

Luo, J., & Isaacowitz, D. M. (2007). How optimists face skin cancer information:

risk assessment, attention, memory, and behavior. Psychology and Health,

22(8), 963-984.

Lyubomirsky, S., Dickerhoof, R., Boehm, J. K., & Sheldon, K. M. (2011). Becoming

happier takes both a will and a proper way: An experimental longitudinal

intervention to boost well-being. Emotion, 11(2), 391.

Mäkikangas, A., & Kinnunen, U. (2003). Psychosocial work stressors and well-being:

Self-esteem and optimism as moderators in a one-year longitudinal sample.


MacLeod, A. K., & Conway, C. (2005). Well‐being and the anticipation of future

positive experiences: The role of income, social networks, and planning

ability. Cognition & Emotion, 19(3), 357-374.

Maier, & Seligman. (1976). Learned helplessness: Theory and evidence. Journal of

Experimental Psychology: General; Journal of Experimental Psychology:

General, 105(1), 3.

Malle, B. F. (2006). The actor-observer asymmetry in attribution: a (surprising)

meta-analysis. Psychological bulletin, 132(6), 895.

Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for

cognition, emotion, and motivation. Psychological review, 98(2), 224.

Marshall, Wortman, C. B., Kusulas, J. W., Hervig, L. K., & Vickers Jr, R. R. (1992).

Distinguishing optimism from pessimism: Relations to fundamental

dimensions of mood and personality. Journal of personality and social


Marshall, M., & Brown, J. (2006). Emotional reactions to achievement outcomes: Is

it really best to expect the worst? Cognition & Emotion, 20(1), 43-63.

Matthews, K. A., Räikkönen, K., Sutton-Tyrrell, K., & Kuller, L. H. (2004).

Optimistic attitudes protect against progression of carotid atherosclerosis in

healthy middle-aged women. Psychosomatic Medicine, 66(5), 640-644.

McCrae, R. R., & Costa, P. T. (1987). Validation of the 5-factor model of personality

across instruments and observers. Journal of personality and social

psychology, 52(1), 81-90. doi: 10.1037/0022-3514.52.1.81

McPherson, J., & Mohr, P. (2005). The role of item extremity in the emergence of

keying-related factors: an exploration with the life orientation test.

Psychological methods, 10(1), 120.

Medway, F. J., & Lowe, C. A. (1976). Effect of stimulus person valence on divergent

self-other attributions for success and failure. Journal of Research in

Personality, 10(3), 266-278. doi: 10.1016/0092-6566(76)90017-9

Meevissen, Y., Peters, M. L., & Alberts, H. J. E. M. (2011). Become more optimistic

by imagining a best possible self: Effects of a two week intervention. Journal

of Behavior Therapy and Experimental Psychiatry.

Metalsky, G. I., Joiner, T. E., Hardin, T. S., & Abramson, L. Y. (1993). Depressive

reactions to failure in a naturalistic setting: A test of the hopelessness and

self-esteem theories of depression. Journal of Abnormal Psychology, 102(1),

101.

Mezulis, A. H., Abramson, L. Y., Hyde, J. S., & Hankin, B. L. (2004). Is there a

universal positivity bias in attributions? A meta-analytic review of individual,


270

developmental, and cultural differences in the self-serving attributional bias.

Psychological bulletin, 130(5), 711.

Michalos, A. C. (1988). Optimism in thirty countries over a decade. Social Indicators

Research, 20(2), 177-180.

Miller, D. T., & Ross, M. (1975). Self-serving biases in the attribution of causality:

Fact or fiction? Psychological bulletin, 82(2), 213.

Miller, J. G. (1984). Culture and the development of everyday social explanation.

Journal of personality and social psychology, 46(5), 961.

Minton, M. E., Hertzog, M., Barron, C. R., French, J. A., & Reiter-Palmon, R. (2009).

The First Anniversary: Stress, Well-Being, and Optimism in Older Widows.

Western Journal of Nursing Research, 31(8), 1035-1056.

Monzani, D., Steca, P., & Greco, A. (2014). Brief report: Assessing dispositional

optimism in adolescence--factor structure and concurrent validity of the Life

Orientation Test--Revised. Journal of adolescence, 37(2), 97-101. doi:

10.1016/j.adolescence.2013.11.006

Mosing, M. A., Medland, S. E., McRae, A., Landers, J. G., Wright, M. J., & Martin,

N. G. (2012). Genetic Influences on Life Span and Its Relationship to

Personality: A 16-Year Follow-Up Study of a Sample of Aging Twins.

Psychosomatic Medicine, 74(1), 16-22. doi:

10.1097/PSY.0b013e3182385784

Mosing, M. A., Pedersen, N. L., Martin, N. G., & Wright, M. J. (2010). Sex

Differences in the Genetic Architecture of Optimism and Health and Their

Interrelation: A Study of Australian and Swedish Twins. Twin Research and

Human Genetics, 13(4), 322-329.

Mosing, M. A., Zietsch, B. P., Shekar, S. N., Wright, M. J., & Martin, N. G. (2009).

Genetic and Environmental Influences on Optimism and its Relationship to

Mental and Self-Rated Health: A Study of Aging Twins. Behav Genet, 39(6),

597-604. doi: 10.1007/s10519-009-9287-7

Musgrave-Marquart, D., Bromley, S. P., & Dalley, M. B. (1997). Personality,

academic attribution, and substance use as predictors of academic

achievement in college students. Journal of Social Behavior and Personality,

12(2), 501-511.

Myers, D. G., & Diener, E. (1995). Who is happy? Psychological Science, 6(1), 10-

19.

Negovan, V., & Bagana, E. (2011). A comparison of relationship between self

esteem and vulnerability to depression among high school and freshmen

university students. In D. E. Ongen, C. Hursen, M. Halat & H. Boz (Eds.),

2nd World Conference on Psychology, Counselling and Guidance-2011 (Vol.

30).

Nes, & Segerstrom. (2006). Dispositional optimism and coping: A meta-analytic

review. Personality and Social Psychology Review, 10(3), 235-251. doi:

10.1207/s15327957pspr1003_3

Nes, L. S., Evans, D. R., & Segerstrom, S. C. (2009). Optimism and College

Retention: Mediation by Motivation, Performance, and Adjustment(1).

Journal of Applied Social Psychology, 39(8), 1887-1912.


271

Nes, L. S., Segerstrom, S. C., & Sephton, S. E. (2005). Engagement and arousal:

Optimism’s effects during a brief stressor. Personality and Social Psychology

Bulletin, 31(1), 111-120.

Nicholls, A. R., Polman, R. C. J., Levy, A. R., & Backhouse, S. H. (2008). Mental

toughness, optimism, pessimism, and coping among athletes. Personality and


Norem, J. K., & Chang, E. C. (2002). The positive psychology of negative thinking.

Journal of Clinical Psychology, 58(9), 993-1001.

Norvell, N. K. (1992). Health and optimism - new research on the relationship

between positive thinking and physical well-being. Journal of Nervous and

Mental Disease, 180(3), 214-214.

Nurmi, J. E. (1992). Cross-cultural differences in self-serving bias - responses to the

Attributional Sytle Questionnaire by American and Finnish students. Journal

of Social Psychology, 132(1), 69-76.

Okolo, C. M. (1992). The effects of computer-based attribution retraining on the

attributions, persistence, and mathematics computation of students with

learning disabilities. Journal of Learning Disabilities, 25(5), 327-334.

Organization, W. H. (2012). Fact sheet of depression.

Park, N., Peterson, & Seligman. (2004). Strengths of character and well-being.

Journal of Social and Clinical Psychology, 23(5), 603-619.

Pearson, R. M., Heron, J., Button, K., Bentall, R. P., Fernyhough, C., Mahedy, L., . . .

Lewis, G. (2015). Cognitive styles and future depressed mood in early

adulthood: The importance of global attributions. Journal of affective

disorders, 171, 60-67. doi: 10.1016/j.jad.2014.08.057

Perry, R. P., Hechter, F. J., Menec, V. H., & Weinberg, L. E. (1993). Enhancing

achievement motivation and performance in college students: An attributional

retraining perspective. Research in Higher Education, 34(6), 687-723.

Perry, R. P., & Penner, K. S. (1990). Enhancing academic achievement in college

students through attributional retraining and instruction. Journal of

educational psychology, 82(2), 262.

Peters, Flink, I. K., Boersma, K., & Linton, S. J. (2010). Manipulating optimism: Can

imagining a best possible self be used to increase positive future expectancies?

The Journal of Positive Psychology, 5(3), 204-211.

Peters, Meevissen, Y. M. C., & Hanssen, M. M. (2013). Specificity of the Best

Possible Self intervention for increasing optimism: Comparison with a

gratitude intervention. Terapia Psicologica, 31(1), 93-100.

Peterson. (1988). Explanatory style as a risk factor for illness. Cognitive therapy and

research, 12(2), 119-132.

Peterson. (2000a). The future of optimism. American Psychologist, 55(1), 44.

Peterson. (2000b). Optimistic explanatory style and health (Vol. 2).

Peterson, & Barrett, L. C. (1987). Explanatory style and academic performance

among university freshmen. Journal of personality and social psychology,

53(3), 603-607. doi: 10.1037/0022-3514.53.3.603

Peterson, Berres, B. A., & Seligman. (1985). Depressive symptoms and unprompted

causal attributions: Content analysis. Behaviour Research and Therapy, 23(4),

379-382.


272

Peterson, & Bossio, L. M. (2001). Optimism and physical well-being. In E. C. Chang

(Ed.), Optimism & pessimism: implications for theory, research, and practice.

Washington, DC: American Psychological Association.

Peterson, & Seligman. (1984). Causal explanations as a risk factor for depression:

Theory and evidence. Psychological review, 91(3), 347.

Peterson, Semmel, A., von Baeyer, C., Abramson, Metalsky, G. I., & Seligman.

(1982). The attributional style questionnaire. Cognitive therapy and research,

6(3), 287-299.

Peterson, & Vaidya, R. S. (2001). Explanatory style, expectations, and depressive

symptoms. Personality and Individual Differences, 31(7), 1217-1223.

Peterson, & Villanova, P. (1988). An Expanded Attributional Style Questionnaire.

Journal of Abnormal Psychology, 97(1), 87.

Peterson, C., & Steen, T. A. (2009). Optimistic Explanatory Style. In S. J. L. C. R.

Snyder (Ed.), Oxford handbook of positive psychology (Vol. 2, pp. 313-321):

Oxford University Press.

Plomin, R., Scheier, M. F., Bergeman, C. S., Pedersen, N. L., Nesselroade, J. R., &

McClearn, G. E. (1992). Optimism, pessimism and mental health: A

twin/adoption analysis. Personality and Individual Differences, 13(8), 921-

930. doi: http://dx.doi.org/10.1016/0191-8869(92)90009-E

Poropat, A. (2002). The relationship between attributional style, gender and the Five-

Factor Model of personality. Personality and Individual Differences, 33(7),

1185-1201. doi: 10.1016/s0191-8869(02)00008-9

Prati, G., & Pietrantoni, L. (2009). Optimism, Social Support, and Coping Strategies

As Factors Contributing to Posttraumatic Growth: A Meta-Analysis. Journal

of Loss & Trauma, 14(5), 364-388. doi: 10.1080/15325020902724271

Pretzer, J. L., & Walsh, C. A. (2001). Optimism, pessimism, and psychotherapy:

Implications for clinical practice. In Chang (Ed.), Optimism and pessimism.

Implications for theory, research and practice.

R Core Team. (2012). R: A language and environment for statistical computing

(Version 2.15.X). Vienna, Austria: R Foundation for Statistical Computing.

Retrieved from http://www.R-project.org

Räikkönen, K., & Matthews, K. A. (2008). Do dispositional pessimism and optimism

predict ambulatory blood pressure during schooldays and nights in

adolescents? Journal of Personality, 76(3), 605-630.

Rand, K. L., Martin, A. D., & Shea, A. M. (2011). Hope, but not optimism, predicts

academic performance of law students beyond previous academic

achievement. Journal of Research in Personality, 45(6), 683-686.

Rasmussen, H. N., Scheier, M. F., & Greenhouse, J. B. (2009). Optimism and

physical health: A meta-analytic review. Annals of Behavioral Medicine,

37(3), 239-256.

Rauch, S. A. M., Defever, E., Oetting, S., Graham-Bermann, S. A., & Seng, J. S.

(2013). Optimism, Coping, and Posttraumatic Stress Severity in Women in

the Childbearing Year. Psychological Trauma-Theory Research Practice and

Policy, 5(1), 77-83. doi: 10.1037/a0022214

Reed, G. L., & Enright, R. D. (2006). The effects of forgiveness therapy on

depression, anxiety, and posttraumatic stress for women after spousal

emotional abuse. Journal of Consulting and Clinical Psychology, 74(5), 920.

http://dx.doi.org/10.1016/0191-8869(92)90009-E

http://www.r-project.org/


273

Reisbig, A. M. J., Danielson, J. A., Wu, T.-F., Hafen, M., Jr., Krienert, A., Girard, D.,

& Garlock, J. (2012). A Study of Depression and Anxiety, General Health,

and Academic Performance in Three Cohorts of Veterinary Medical Students

across the First Three Semesters of Veterinary School. Journal of Veterinary

Medical Education, 39(4), 341-358. doi: 10.3138/jvme.0712-065R

Rhemtulla, M., Brosseau-Liard, P. E., & Savalei, V. (2012). When can categorical

variables be treated as continuous? A comparison of robust continuous and

categorical SEM estimation methods under suboptimal conditions. Psychol

Methods, 17(3), 354-373. doi: 10.1037/a0029315

Rim, Y. (1991). Personality and attributional styles. Personality and Individual

Differences, 12(1), 95-96. doi: 10.1016/0191-8869(91)90137-z

Riskind, J. H., Sarampote, C. S., & Mercier, M. A. (1996). For every malady a

sovereign cure: Optimism training. Journal of Cognitive Psychotherapy,

10(2), 105-117.

Rius-Ottenheim, N., Kromhout, D., de Craen, A. J., Geleijnse, J. M., van der Mast, R.

C., Zitman, F. G., . . . Giltay, E. J. (2012). Parental longevity correlates with

offspring’s optimism in two cohorts of community-dwelling older subjects.

Age, 34(2), 461-468.

Rogers, M. E., Hansen, N. B., Levy, B. R., Tate, D. C., & Sikkema, K. J. (2005).

Optimism and coping with loss in bereaved HIV-infected men and women.

Journal of Social and Clinical Psychology, 24(3), 341-360. doi:

10.1521/jscp.24.3.341.65619

Rotter, J. B. (1966). Generalized expectancies for internal versus external control of

reinforcement. Psychol Monogr, 80(1), 1-28.

Roysamb, E., & Strype, J. (2002). Optimism and pessimism: Underlying structure

and dimensionality. Journal of Social and Clinical Psychology, 21(1), 1-19.

Rupke, S. J., Blecke, D., & Renfrow, M. (2006). Cognitive therapy for depression.

Am Fam Physician, 73(1), 82-86.

Ruthig, J. C., Perry, R. P., Hall, N. C., & Hladkyj, S. (2004). Optimism and

Attributional Retraining: Longitudinal Effects on Academic Achievement,

Test Anxiety, and Voluntary Course Withdrawal in College Students1.

Journal of Applied Social Psychology, 34(4), 709-730.

Ryff. (1989). Happiness is everything, or is it? Explorations on the meaning of

psychological well-being. Journal of personality and social psychology,

57(6), 1069.

Ryff. (1995). Psychological well-being in adult life. Current directions in

psychological science, 4(4), 99-104.

Ryff, & Keyes, C. L. M. (1995). The structure of psychological well-being revisited.

Journal of personality and social psychology, 69(4), 719-727. doi:

10.1037/0022-3514.69.4.719

Ryff, & Singer, B. H. (2006). Know Thyself and Become What You Are: A

Eudaimonic Approach to Psychological Well-Being. Journal of Happiness

Studies, 9(1), 13-39. doi: 10.1007/s10902-006-9019-0

Sanjuán, P., Arranz, H., & Castro, A. (2012). Pessimistic attributions and coping

strategies as predictors of depressive symptoms in people with coronary heart

disease. Journal of health psychology, 17(6), 886-895.


274

Sanjuan, P., & Magallares, A. (2014). Coping Strategies as Mediating Variables

Between Self-serving Attributional Bias and Subjective Well-Being. Journal

of Happiness Studies, 15(2), 443-453. doi: 10.1007/s10902-013-9430-2

Scheier, & Carver. (1985). Optimism, coping, and health: assessment and

implications of generalized outcome expectancies. Health Psychology, 4(3),

219.

Scheier, & Carver. (1987). Dispositional optimism and physical well‐being: the

Influence of generalized outcome expectancies on health. Journal of

Personality, 55(2), 169-210.

Scheier, & Carver. (1992). Effects of optimism on psychological and physical well-

being: Theoretical overview and empirical update. Cognitive therapy and

research, 16(2), 201-228.

Scheier, & Carver. (1993). On the power of positive thinking: The benefits of being

optimistic. Current directions in psychological science, 2(1), 26-30.

Scheier, Carver, & Bridges. (1994). Distinguishing optimism from neuroticism (and

trait anxiety, self-mastery, and self-esteem): A reevaluation of the Life

Orientation Test. Journal of personality and social psychology, 67(6), 1063.

Scheier, Weintraub, J. K., & Carver. (1986). Coping with stress: divergent strategies

of optimists and pessimists. Journal of personality and social psychology,

51(6), 1257.

Scheier., Carver, C. S., & Bridges, M. W. (2001). Optimism, pessimism, and

psychological well-being (chapter 9). Chang, Edward C. (Ed), (2001).

Optimism & pessimism: Implications for theory, research, and practice.

Schou, I., Ekeberg, O., & Ruland, C. M. (2005). The mediating role of appraisal and

coping in the relationship between optimism-pessimism and quality of life.

Psycho-Oncology, 14(9), 718-727. doi: 10.1002/896

Schulman, P., Keith, D., & Seligman. (1993). Is optimism heritable? A study of

twins. Behaviour Research and Therapy, 31(6), 569-574. doi: 10.1016/0005-

7967(93)90108-7

Segerstrom, S. C. (2007). Optimism and resources: Effects on each other and on

health over 10 years. Journal of Research in Personality, 41(4), 772-786. doi:

10.1016/j.jrp.2006.09.004

Segerstrom, S. C., Castañeda, J. O., & Spencer, T. E. (2003). Optimism effects on

cellular immunity: testing the affective and persistence models. Personality

and Individual Differences, 35(7), 1615-1624.

Segerstrom, S. C., Evans, D. R., & Eisenlohr-Moul, T. A. (2011). Optimism and

pessimism dimensions in the Life Orientation Test-Revised: Method and

meaning. Journal of Research in Personality, 45(1), 126-129. doi:

10.1016/j.jrp.2010.11.007

Segerstrom, S. C., & Sephton, S. E. (2010). Optimistic Expectancies and Cell-

Mediated Immunity: The Role of Positive Affect. Psychological Science,

21(3), 448-455. doi: 10.1177/0956797610362061

Seligman. (1989). Explanatory style - predicting depression, achievement, and health.

Seligman. (1991). Learned optimism: New York: Knopf.

Seligman. (2002). Positive psychology, positive prevention, and positive therapy.

Handbook of positive psychology, 2, 3-12.


275

Seligman. (2011). Learned optimism: How to change your mind and your life:

William Heinemann.

Seligman, Abramson, L. Y., Semmel, A., & Von Baeyer, C. (1979). Depressive

attributional style. Journal of Abnormal Psychology, 88(3), 242.

Seligman, & Csikszentmihalyi, M. (2000). Positive psychology: An introduction.

American Psychologist; American Psychologist, 55(1), 5.

Seligman, Rashid, T., & Parks, A. C. (2006). Positive psychotherapy. American

Psychologist, 61(8), 774.

Seligman, & Schulman, P. (1986). Explanatory style as a predictor of productivity

and quitting among life insurance sales agents. Journal of personality and


Seligman, Steen, T. A., Park, N., & Peterson. (2005). Positive psychology progress:

empirical validation of interventions. American Psychologist, 60(5), 410.

Sergeant, S., & Mongrain, M. (2014). An Online Optimism Intervention Reduces

Depression in Pessimistic Individuals. Journal of Consulting and Clinical

Psychology, 82(2), 263-274. doi: 10.1037/a0035536

Shapcott, K. M., & Carron, A. V. (2010). Development and validation of a team

attributional style questionnaire. Group Dynamics: Theory, Research, and

Practice, 14(2), 93.

Shapira, L. B., & Mongrain, M. (2010). The benefits of self-compassion and

optimism exercises for individuals vulnerable to depression. The Journal of

Positive Psychology, 5(5), 377-389.

Sharifi, M., Hajiheidari, M., Khorvash, F., & Mirabdollahi, M. A. (2013).

Effectiveness of Attribution Retraining on Women's Depression and Anxiety

After Miscarriage. International journal of preventive medicine, 4(Suppl 2),

S239-244.

Sharot, T., Riccardi, A. M., Raio, C. M., & Phelps, E. A. (2007). Neural mechanisms

mediating optimism bias. Nature, 450(7166), 102-105.

Sharpe, J. P., Martin, N. R., & Roth, K. A. (2011). Optimism and the Big Five

factors of personality: Beyond Neuroticism and Extraversion. Personality and


Shek, D. T. (1990). Reliability and factorial structure of the Chinese version of the

Beck Depression Inventory. Journal of Clinical Psychology, 46(1), 35-43.

Sheldon, K. M., & Lyubomirsky, S. (2006). How to increase and sustain positive

emotion: The effects of expressing gratitude and visualizing best possible

selves. The Journal of Positive Psychology, 1(2), 73-82.

Sin, N. L., Della Porta, M. D., & Lyubomirsky, S. (2011). Tailoring Positive

Psychology Interventions to Treat Depressed Individuals.

Sin, N. L., & Lyubomirsky, S. (2009). Enhancing Well-Being and Alleviating

Depressive Symptoms With Positive Psychology Interventions: A Practice-

Friendly Meta-Analysis. Journal of Clinical Psychology, 65(5), 467-487. doi:

10.1002/jclp.20593

Sinha, B. K., Willson, L. R., & Watson, D. C. (2000). Stress and coping among

students in India and Canada. Canadian Journal of Behavioural

Science/Revue canadienne des sciences du comportement, 32(4), 218.


276

Smith, P., Caputi, P., & Crittenden, N. (2013). Measuring optimism in organizations:

development of a workplace explanatory style questionnaire. Journal of

Happiness Studies, 14(2), 415-432.

Snyder. (1989). Reality negotiation: From excuses to hope and beyond. Journal of

Social and Clinical Psychology, 8(2), 130-157. doi:

10.1521/jscp.1989.8.2.130

Snyder. (2002). Hope theory: Rainbows in the mind. Psychological Inquiry, 13(4),

249-275.

Snyder, Harris, C., Anderson, J. R., Holleran, S. A., Irving, L. M., Sigmon, S. T., . . .

Harney, P. (1991). The will and the ways: development and validation of an

individual-differences measure of hope. Journal of personality and social


Snyder, Sympson, S. C., Ybasco, F. C., Borders, T. F., Babyak, M. A., & Higgins, R.

L. (1996). Development and validation of the State Hope Scale. Journal of

personality and social psychology, 70(2), 321-335. doi: 10.1037//0022-

3514.70.2.321

Stanton, A. L., Danoff-Burg, S., Sworowski, L. A., Collins, C. A., Branstetter, A. D.,

Rodriguez-Hanley, A., . . . Austenfeld, J. L. (2002). Randomized, controlled

trial of written emotional expression and benefit finding in breast cancer

patients. Journal of Clinical Oncology, 20(20), 4160-4168.

Steed, L. G. (2002). A psychometric comparison of four measures of hope and

optimism. Educational and Psychological Measurement, 62(3), 466-482.

Strutton, D., & Lumpkin, J. (1992). Relationship between optimism and coping

strategies in the work environment. Psychological Reports, 71(3), 1179-1186.

doi: 10.2466/pr0.71.8.1179-1186

Suls, J., & Fletcher, B. (1985). The relative efficacy of avoidant and nonavoidant

coping strategies: A meta-analysis. Health Psychology, 4(3), 249-288. doi:

10.1037/0278-6133.4.3.249

Sumi, K. (2004). The Japanese version of the revised Life orientation Test:

Reliability and construct validity. Psychological Reports, 95(1), 86-88. doi:

10.2466/pr0.95.1.86-88

Sweeney, P. D., Anderson, K., & Bailey, S. (1986). Attributional style in depression:

A meta-analytic review. Journal of personality and social psychology, 50(5),

974-991. doi: 10.1037/0022-3514.50.5.974

Taylor, S. E., Kemeny, M. E., Aspinwall, L. G., Schneider, S. G., Rodriguez, R., &

Herbert, M. (1992). Opitmism, coping, psychological distress, and high-risk

sexual-behavior among men at risk for acquired-immunodeficiency-

syndrome (AIDS). Journal of personality and social psychology, 63(3), 460-

473. doi: 10.1037//0022-3514.63.3.460

Teglasi, H., & Fagin, S. S. (1984). Social anxiety and self-other biases in causal

attribution. Journal of Research in Personality, 18(1), 64-80.

Thompson, A. S., Gaudreau, P., Hoar, S. D., Hadd, V., & Lelievre, J. (2008).

Domain specific coping as intermediate processes in the relationships

between dispositional optimism-pessimism and satisfaction in the school and

sport domains. Journal of Sport & Exercise Psychology, 30, S205-S205.

Tiger, L. (1979). Optimism: The biology of hope (Vol. 1st Touchstone edition). New

York: Simon and Schuster.


277

Tiggemann, M., Winefield, A. H., Winefield, H. R., & Goldney, R. D. (1991). The

prediction of psychological distress from attributional style: A test of the

hopelessness model of depression. Australian journal of psychology, 43(3),

125-127.

Tindle, H. A., Chang, Y.-F., Kuller, L. H., Manson, J. E., Robinson, J. G., Rosal, M.

C., . . . Matthews, K. A. (2009). Optimism, cynical hostility, and incident

coronary heart disease and mortality in the Women’s Health Initiative.

Circulation, 120(8), 656-662.

Tomakowsky, J., Lumley, M. A., Markowitz, N., & Frank, C. (2001). Optimistic

explanatory style and dispositional optimism in HIV-infected men. Journal of

Psychosomatic Research, 51(4), 577-587. doi: 10.1016/s0022-

3999(01)00249-5

Tseng, W. C. (2007). Social Support as a Mediator of Optimism and Well-being in

College Students: A Structural Equation Modeling Approach (Article written

in Chinese). Journal of Education and Psychology, 30(4), 117-146.

Vacek, K. R., Coyle, L. D., & Vera, E. M. (2010). Stress, Self-Esteem, Hope,

Optimism, and Well-Being in Urban, Ethnic Minority Adolescents. Journal

of Multicultural Counseling and Development, 38(2), 13.

Vazquez, C., Jimenez, F., Saura, F., & Avia, M. D. (2001). The importance of

"importance": a longitudinal confirmation of the attributional-stress model of

depression measuring the importance of the attributions and the impact of the

stressor. Personality and Individual Differences, 31(2), 205-214. doi:

10.1016/s0191-8869(00)00129-x

Wang, C., & Zhang, N. (2005). Personality Correlates to Attributional Style in

Undergraduates. Chinese Journal of Clinical Psychology, 13(1), 53-54.

Watson, D. (1982). The actor and the observer: How are their perceptions of

causality divergent? Psychological bulletin, 92(3), 682.

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of

brief measures of positive and negative affect: the PANAS scales. Journal of

personality and social psychology, 54(6), 1063.

Weiner, B. (1974). Achievement motivation and attribution theory: General Learning

Press.

Weiner, B. (1985). An attributional theory of achievement motivation and emotion.

Psychological review, 92(4), 548.

Wiffen, B. D. R., O'Connor, J. A., Gayer-Anderson, C., Marques, T. R., McQueen,

G., Happe, F., . . . David, A. S. (2013). "I am sane but he is mad": Insight and

illness attributions to self and others in psychosis. Psychiatry Research,

207(3), 173-178. doi: 10.1016/j.psychres.2013.01.020

Williams, D. (1992). Dispositional optimism, neuroticism, and extraversion.


Wise, D., & Rosqvist, J. (2006). Explanatory Style and Well-Being. In D. L. S. Jay C.

Thomas (Ed.), Comprehensive Handbook of Personality and

Psychopathology (pp. 285-305). Newjercy: John Wiley & Sons, Inc.

Wrosch, C., & Scheier, M. F. (2003). Personality and quality of life: The importance

of optimism and goal adjustment. Quality of Life Research, 12, 59-72.

Yang, J., McCrae, R. R., Costa Jr, P. T., Dai, X., Yao, S., Cai, T., & Gao, B. (1999).

Cross-cultural personality assessment in psychiatric populations: The NEO-


278

PI—R in the People's Republic of China. Psychological Assessment, 11(3),

359.

You, J., Fung, H. H., & Isaacowitz, D. M. (2009). Age differences in dispositional

optimism: A cross-cultural study. European Journal of Ageing, 6(4), 247-252.

Yu, & Seligman. (2002). Preventing depressive symptoms in Chinese children.

Prevention & Treatment, 5(1), 9a.

Yu, C. Y. (2002). Evaluating cutoff criteria of model fit indices for latent variable

models with binary and continuous outcomes. University of California, Los

Angeles, Los Angeles. Retrieved from

http://www.statmodel.com/download/Yudissertation.pdf

Yu, J. J., & Ko, Y. K. (2013). Paternal family expressiveness as a mediator between

father's dispositional optimism and child's dispositional optimism. The

Journal of genetic psychology, 174(6), 677-695.

Yuan, L., & Zhang, R. (2007). A Research on the Relationships among Explanatory

style, Dispositional Optimism and Psychological well-being. Psychological

Development and Education, 2, 111-115.

Zhang, C. W. N. (2006). A Study on the Scale of Attributional Style Questionnaire.

China Journal of Behavioral Medical Science, 15(5), 470-471.

http://www.statmodel.com/download/Yudissertation.pdf

Liu2016.pdf - Edinburgh Research Archive

Documents